1. No category

Cellular Respiration and Carcinogenesis
Shireesh P. Apte Rangaprasad Sarangarajan
•
Editors
Cellular Respiration
and Carcinogenesis
Foreword by Gregg L. Semenza
123
Editors
Shireesh P. Apte
Alcon Laboratories
Fort Worth, TX
USA
shireesh.apte@alconlabs.com
ISBN: 978-1-934115-07-7
DOI 10.1007/978-1-59745-435-3
Rangaprasad Sarangarajan
Massachusetts College of
Pharmacy and Health Sciences
Worcester, MA
USA
rangaprasad.sarangarajan@mcphs.edu
e-ISBN: 978-1-59745-435-3
Library of Congress Control Number: 2008937240
c Humana Press, a part of Springer Science+Business Media, LLC 2009
All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Humana Press, c/o Springer Science+Business Media, LLC, 233 Spring Street,
New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis.
Use in connection with any form of information storage and retrieval, electronic adaptation, computer
software, or by similar or dissimilar methodology now known or hereafter developed is forbidden.
The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are
not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject
to proprietary rights.
Printed on acid-free paper
springer.com
Foreword
Cancer is remarkably heterogeneous when one considers pathogenesis of the disease
in affected individuals with the same diagnosis (such as “renal carcinoma”) or even
when cancer in a single individual is considered over time and space. This heterogeneity is the single greatest obstacle to effective cancer therapy. As a result of
disease heterogeneity, there are not too many universal truths in cancer biology. One
of the most consistent findings, though, is the dramatic increase in glucose uptake
by metastatic cancer cells compared with that of their normal counterparts. This
is such a reliable characteristic that it is used clinically to identify occult disease
by positron emission tomography after injection of [18 F]fluoro-2-deoxy-D-glucose.
Increased glucose uptake reflects an increased rate of glycolysis, with conversion of
glucose to lactate and decreased conversion of pyruvate to acetyl coenzyme A, the
substrate for mitochondrial oxidative phosphorylation. Because of the relative inefficiency of glycolysis compared with that of oxidative phosphorylation as a means
of generating ATP, flux through the glycolytic pathway must increase dramatically
to maintain cellular energetics.
Because it is observed in more than 95% of advanced cancers, understanding
the mechanisms and consequences of this dramatic reprogramming of glucose and
energy metabolism in cancer cells is an important challenge for contemporary cancer biology. The 12 chapters of this book provide the reader with state-of-the-art
summaries of recent exciting progress that has been made in the field of cancer
metabolism. This story starts at the beginning of the 20th century with the work
of Otto Warburg, whom Hans Krebs described as a “Cell Physiologist, Biochemist,
and Eccentric.” Warburg invented a device, now known as the Warburg manometer,
with which he demonstrated that tumor cells consume less O2 (and produce more
lactate) than do normal cells under the same ambient O2 concentrations. A century
later, the struggle to understand how and why metastatic cancer cells manifest the
Warburg effect is still ongoing, and 12 rounds of this heavyweight fight await the
reader beyond this brief introduction.
Why do cancer cells manifest altered energy metabolism? To answer this question, it is necessary to acknowledge that mitochondria, in addition to serving as the
major source of ATP production, provide two other critical cellular functions. One
is that they also serve as the major source of reactive oxygen species (ROS), which
are generated when electrons donated to the respiratory chain are not successfully
v
vi
Foreword
transferred from complex IV (cytochrome c oxidase) to O2 to form water but instead
react with O2 prematurely to form superoxide anion, which is then converted to
hydrogen peroxide through the action of mitochondrial superoxide dismutase. ROS
are used by normal cells as signaling molecules (e.g., during cell proliferation).
However, if generated in excess (amplitude or duration or both), ROS result in the
oxidation of lipids, proteins, and nucleic acids leading to cell dysfunction or death,
which introduces the other major role of mitochondria as arbiters of cell survival,
as apoptosis is triggered through changes in mitochondrial structure and function.
Thus, it appears that reduced mitochondrial respiration, which often is accompanied
by reduced mitochondrial mass, occurs in cancer cells as a result of selection against
excess ROS production and/or apoptosis.
How do cancer cells reprogram energy metabolism? To answer this question
(and to more completely answer “why?” beyond the pat generalization provided
above), it is necessary to return full circle to the issue of heterogeneity. Because
mitochondria play critical roles in energy production, redox balance, proliferation,
and survival, many mutations that are selected for in advanced cancers impact mitochondrial function in one or more ways, for although mitochondria control cell fate,
the cell controls mitochondria by virtue of the fact that most of the components of
mitochondria are the products of nuclear, not mitochondrial, genes.
A danger of a reductionist approach to cancer biology that focuses on changes
involving key oncogenes and tumor suppressor genes is the misconception that carcinogenesis can be attributed to one or a few mutations in any given cell. As in
every other aspect of cancer biology, it is the net effect of all of the mutations in
nuclear and mitochondrial DNA of the cancer cell that will determine its metabolic
phenotype. Another danger of a solely mutation-based approach to cancer is that it
ignores the effect of the tumor microenvironment. Like normal cells, cancer cells
react to changes in the concentrations of O2 , glucose, hydrogen ions, and free radicals. In fact, these changes represent the most important selective force acting on the
mutant gene pool of cancer cells. As a result, not all changes in cancer metabolism
are hard-wired by genetic alteration but may instead represent adaptive responses to
the microenvironment. In other words, the reprogramming of energy metabolism is
unique to each cancer cell.
The complex considerations that are covered so superficially in this brief introduction are discussed in detail in the 12 chapters that follow. There is tremendous
interest (and hope) that because altered glucose and energy metabolism appears to
be a virtually universal characteristic of metastatic cancer cells (i.e., the cells that
kill cancer patients), understanding the mechanisms and consequences of metabolic
reprogramming may lead to novel therapeutic approaches that will more effectively
fight this formidable foe. Ring the bell for Round 1 . . .
Baltimore, Maryland
Gregg L. Semenza, M.D., Ph.D.
Contents
Oxidative Phosphorylation and Cancer: The Ongoing
Warburg Hypothesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Michael Ristow and Jos´e M. Cuezva
1
The Electron Transport Chain and Carcinogenesis . . . . . . . . . . . . . . . . . . . . . 19
Jean-Jacques Bri`ere, Paule B´enit, and Pierre Rustin
Respiratory Control of Redox Signaling and Cancer . . . . . . . . . . . . . . . . . . . 33
Pauline M. Carrico, Nadine Hempel, and J. Andr´es Melendez
Cellular Respiration and Dedifferentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Roberto Scatena, Patrizia Bottoni, and Bruno Giardina
Cellular Adaptations to Oxidative Phosphorylation Defects in Cancer . . . . 55
Sarika Srivastava and Carlos T. Moraes
Regulation of Glucose and Energy Metabolism in Cancer Cells by
Hypoxia Inducible Factor 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
´
T´ulio C´esar Ferreira and Elida
Geralda Campos
The Role of Glycolysis in Cellular Immortalization . . . . . . . . . . . . . . . . . . . . . 91
Hiroshi Kondoh
Metabolic Modulation of Carcinogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Shireesh P. Apte and Rangaprasad Sarangarajan
Mitochondrial DNA Mutations in Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Anna Czarnecka and Ewa Bartnik
Cellular Respiration and Tumor Suppressor Genes . . . . . . . . . . . . . . . . . . . . . 131
Luis F. Gonzalez-Cuyar, Fabio Tavora, Iusta Caminha, George Perry,
Mark A. Smith, and Rudy J. Castellani
vii
viii
Contents
Uncoupling Cellular Respiration: A Link to Cancer Cell Metabolism
and Immune Privilege . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
M. Karen Newell, Elizabeth M. Villalobos-Menuey, Marilyn Burnett,
and Robert E. Camley
How Cancer Cells Escape Death . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Erica Werner
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Contributors
Shireesh P. Apte
Senior Scientist, Process Research and Support, Mail stop: R4-11 Alcon
Laboratories, 6201 South Freeway, Fort Worth, Texas 76134, USA
Ewa Bartnik
Department of Genetics and Biotechnology, University of Warsaw; and
Institute of Biochemistry and Biophysics, Polish Academy of Sciences Warsaw,
Pawinskiogo 5a. 02106, Poland
Paule B´enit
Hˆopital Robert Debr´e, INSERM U676, Paris, France
Patrizia Bottoni
Dipartimento di Medicina di Laboratorio, Universit`a Cattolica, Rome, Italy
Jean-Jacques Bri`ere
Hˆopital Robert Debr´e, INSERM U676, Paris, France
Marilyn Burnett
CU Institute of Bioenergetics and Immunology, University of Colorado at Colorado
Springs, Colorado Springs, Colorado
Iusta Caminha
Department of Pharmacology, University of Maryland, Baltimore, Maryland
Robert E. Camley
CU Institute of Bioenergetics and Immunology, University of Colorado at Colorado
Springs, Colorado Springs, Colorado
´
Elida
Geralda Campos
Departamento de Biologia Celular, Universidade de Bras´ılia, Bras´ılia, DF, Brazil
Pauline M. Carrico
Ordway Research Institute, Albany, New York
Rudy J. Castellani
Department of Pathology, University of Maryland, Baltimore, Maryland
ix
x
Contributors
Jos´e M. Cuezva
Departamento de Biolog´ıa Molecular, Centro de Biolog´ıa Molecular “Severo
Ochoa,” Centro de Investigaci´on Biom´edica en Red de Enfermedades Raras,
Universidad, Aut´onoma de Madrid, Madrid, Spain
Anna Czarnecka
Department of Genetics and Biotechnology, University of Warsaw; Warsaw,
Poland; and Postgraduate School of Molecular Medicine, Warsaw, Poland
T´ulio C´esar Ferreira
Departamento de Biologia Celular, Universidade de Bras´elia, Bras´elia, DF, Brazil
Bruno Giardina
Dipartimento di Medicina di Laboratorio, Universit`a Cattolica, Rome, Italy
Luis F. Gonzalez-Cuyar
Department of Pathology, University of Maryland, Baltimore, Maryland
Nadine Hempel
The Center for Immunology & Microbial Disease, Albany Medical College,
Albany, New York
Hiroshi Kondoh
Department of Geriatric Medicine, Graduate School of Medicine, Kyoto University,
Kyoto, Japan
J. Andr´es Melendez
The Center for Immunology & Microbial Disease, Albany Medical College,
Albany, New York
Carlos T. Moraes
Department of Neurology and Department of Cell Biology and Anatomy,
University of Miami Miller School of Medicine, Miami, Florida
M. Karen Newell
CU Institute of Bioenergetics and Immunology, University of Colorado at Colorado
Springs, Colorado Springs, Colorado
George Perry
College of Sciences, University of Texas at San Antonio, San Antonio, Texas;
and Department of Pathology, Case Western Reserve University, Cleveland, Ohio
Michael Ristow
Department of Human Nutrition, Institute of Nutrition, University of Jena, Jena,
Germany
Pierre Rustin
Hˆopital Robert Debr´e, Paris, France
Contributors
xi
Rangaprasad Sarangarajan
Associate Professor of Pharmacology & Toxicology, School of Pharmacy
(Worcester) Massachusetts College of Pharmacy and Health Sciences,
19 Foster Street, Worcester, MA 01608,
Gregg L. Semenza
Director, Vascular Program, Institute for Cell Engineering; Professor, Departments
of Pediatrics, Medicine, Oncology, Radiation Oncology, and the McKusick-Nathans
Institute of Genetic Medicine, The Johns Hopkins University School of Medicine,
Broadway Research Building, Suite 671, 733 North Broadway, Baltimore, MD
21205 USA
Roberto Scatena
Dipartimento di Medicina di Laboratorio, Universit`a Cattolica del Sacro Cuore,
Rome, Italy
Mark A. Smith
Department of Pathology, Case Western Reserve University, Cleveland, Ohio
Sarika Srivastava
Department of Neurology, University of Miami Miller School of Medicine, Miami,
Florida
Fabio Tavora
Department of Pathology, University of Maryland, Baltimore, Maryland
Elizabeth M. Villalobos-Menuey
CU Institute of Bioenergetics and Immunology, University of Colorado at Colorado
Springs, Colorado Springs, Colorado
Erica Werner
Department of Cell Biology, School of Medicine, Emory University, Atlanta,
Georgia
Oxidative Phosphorylation and Cancer:
The Ongoing Warburg Hypothesis
Michael Ristow and Jos´e M. Cuezva
Abstract More than eight decades ago, the German physiologist Otto Warburg
observed that cancer cells in the presence of oxygen produced large amounts of lactate and proposed that impaired oxidative metabolism may cause cancer. This formulation, later known as the Warburg hypothesis, was investigated and debated for
several decades. The development of molecular biology and the discovery of oncogenes and tumor suppressor genes in subsequent years shifted the general interest in
the cancer field into directions other than metabolism and promoted the abandoning
of the Warburg hypothesis or its consideration as an epiphenomenon of cell transformation. In recent years, a renaissance in the field of mitochondria has occurred
in biological studies. This has happened mostly by the recognition of the key functional role that this organelle plays in the execution of cell death and, thus, for its
participation in the development of a vast array of human pathologies. The cancer
field was not indifferent to these changes, and the Warburg hypothesis was brought
back to the scene with renewed strength. Specifically, recent findings suggest that
cancer is associated with a decrease in the activity and expression of ␤-F1-ATPase,
a key subunit of the mitochondrial ATP synthase. This alteration has been shown to
limit oxidative phosphorylation and to trigger the induction of glycolysis to provide
energy to the cell thus configuring the earlier Warburg observation in an additional
hallmark of the cancer cell. Moreover, increased and decreased cellular mitochondrial activities are respectively associated with suppression and development of cancer. We suggest that reactivating mitochondrial metabolism by pharmacologic or
dietary measures and/or tackling the deviant glycolysis in cancer cells may efficiently suppress malignant growth.
M. Ristow (B)
Chair, Department of Human Nutrition, Institute of Nutrition, University of Jena, D-07743 Jena,
Germany
e-mail: michael.ristow@mristow.org
J.M. Cuezva (B)
Departamento de Biolog´ıa Molecular, Centro de Biolog´ıa Molecular “Severo Ochoa”, CIBER de
Entermedades Rasas, Universidad Aut´onoma de Madrid, 28049 Madrid, Spain.
e-mail: jmcuezva@cbm.uam.es
S.P. Apte, R. Sarangarajan (eds.), Cellular Respiration and Carcinogenesis,
DOI 10.1007/978-1-59745-435-3 1,
C Humana Press, a part of Springer Science+Business Media, LLC 2009
1
2
M. Ristow, J.M. Cuezva
Keywords Mitochondria · Oxidative phosphorylation · H+ -ATP synthase ·
Glycolysis · Cancer
Nowadays, malignant disorders are a major cause of premature death. The incidence
of cancers increases with age, whereas metabolic activity and resting energy expenditure resembling oxidative metabolism continuously decrease during aging of
humans. Hence, an inverse relationship between mitochondrial activity and the promotion of tumor growth might possibly exist. To analyze whether this apparent link
is supported by scientific evidence rather than simply being a coincidence, tremendous efforts have been undertaken in the past to elucidate this possible connection.
This chapter aims to summarize published evidence on mechanistic interdependencies between cancer biology and the activity of mitochondria. Other recent reviews
dealing with this subject are also available [1, 2].
In the past decades, glucose has become a major component of Western diets [3].
The most frequently used nutritive sugar, the disaccharide sucrose, contains equal
amounts of the monosaccharides fructose and glucose. Glucose is actively transported into mammalian cells, and subsequently is (as well as fructose) biochemically converted into pyruvate (Fig. 1). Pyruvate either is converted into lactate,
which prevents its further intracellular breakdown, or is completely oxidized to carbon dioxide and water by mitochondrial activities linked to the supply of molecular
oxygen (Fig. 1). The breakdown of glucose into pyruvate is termed glycolysis.
Glycolysis takes place in the cytoplasm of eukaryotic cells, and it can proceed both
in the absence and presence of oxygen (Fig. 1). In order to carry on glycolysis in
the former case, the carbon skeleton of pyruvate needs to be reduced to lactate to
regenerate the pool of oxidized NAD+ (Fig. 1). Many prokaryotes and lower eukaryotes including Saccharomyces cerevisiae (yeast) survive in the absence of oxygen
by obtaining energy from anaerobic glycolysis only [4], whereas higher eukaryotes,
especially mammals, require a continuous supply of oxygen for obtaining biological energy. Glycolysis in absence or presence of oxygen provides comparably little
energy equivalents per mole of glucose oxidized when compared with its terminal
oxidation in the mitochondria (Fig. 1). Nevertheless, anaerobic glycolysis supplies
the precursors and building blocks for the biosynthesis of cellular macromolecules,
and, in yeast, as well as in certain mammalian cell types, is sufficient to maintain
viability for longer periods of time. Because highly developed organisms tend to
use nutritive energy equivalents more efficiently than do less developed organisms
[5], mammals have significantly more efficient ways to generate energy equivalents.
This is illustrated by the further oxidation of a metabolite of glycolysis (pyruvate) in
mitochondria, which yields an approximately 18-fold amount of energy equivalents
per mole of glucose oxidized when compared with anaerobic glycolysis (Fig. 1).
Hence, the major advantage of glycolytic metabolism is to be independent from
oxygen, and the major disadvantage of this process is a comparably low yield of
energy equivalents (Fig. 1).
For many decades, mitochondrial organelles were considered eminently important for the generation of biological energy and as the cellular site that allowed
Oxidative Phosphorylation and Cancer: The Ongoing Warburg Hypothesis
HK
Glucose
PFK1
PGI
G-6-P
F-6-P
3
ALD
F-1,6-P
TPI
G-3-P
G-3-P + DHAP
GAPDH
ENO
PEP
PGM
2-PG
PGK
3-PG
1,3-BPG
PK
02
Pyruvate
NAD+
LDH
Lactate
Net ATP produced: 2
Glucose utilization: HIGH
AEROBIOSIS
AEROBIOSIS /
ANAEROBIOSIS
NADH
pyruvate
CO2 + H2O
Net ATP produced: 36
Glucose utilization: LOW
Fig. 1 Cellular ATP provision defines the rate of aerobic glycolysis. Abbreviations of glycolytic
intermediates and enzymes of glycolysis are presented sequentially up to the pyruvate crossroad.
The key glycolytic enzymes hexokinase (HK), phosphofructokinase 1 (PFK1), and pyruvate kinase
(PK) are highlighted in bold. The reduction of pyruvate to lactate by lactate dehydrogenase (LDH)
regenerates NAD+ and allows glycolysis to proceed with very low yield of energy and a high
consumption of glucose. Pyruvate could be oxidized to CO2 and water only in the presence of
oxygen by mitochondria. In this situation, the yield of energy obtained is high and consequently
glucose consumption is low
the integration of the metabolism of macronutrients, specifically carbohydrates,
fatty acids, and amino acids. Interconversion of carbohydrate derivatives (i.e., pyruvate) into fatty acids, amino acids into carbohydrate precursors, and others occur in
the matrix of mitochondria, and these activities are integrated in the Krebs cycle.
Concurrently, the reducing equivalents obtained from the terminal oxidation of the
metabolic intermediates of the Krebs cycle in the form of NADH and FADH2 are
shuttled into the complexes of the respiratory chain (Fig. 2). The respiratory chain
is composed of proteins that are located in the inner mitochondrial membrane and
have, as overall function, the reoxidation of NADH and FADH2 to further allow the
oxidation of more metabolic substrates (Fig. 2). The ordered transfer of electrons
along the carriers of the respiratory chain to molecular oxygen (Fig. 2), which is
the final electron acceptor, is the process known as respiration. Electron transfer
in complexes of the respiratory chain trigger the proton pumping activity of complexes CI, CIII, and CIV (Fig. 2), which extrudes protons from the matrix interior of
mitochondria and finally results in the generation of the proton electrochemical gradient (⌬␮H+ ) across the inner membrane of the organelle, which is the intermediate
used for the synthesis of ATP [6] (Fig. 2). The reentry of protons to the matrix interior through the proton channel of the H+ -ATP synthase (CV in Fig. 2) drives the
4
M. Ristow, J.M. Cuezva
O2·
CYTOPLASM
O.M.
H+
H+
Cyt c
–
CI e
Q
NAD+
FADH2
CIV
e–
I.M.
CIII
–
CII e
NADH
ΔμH+
H+
e–
FAD
O2 H2O
CV
ATP
H+
MATRIX
O2·
Fig. 2 Mitochondrial activities: respiration, oxidative phosphorylation, and the production of ROS.
The oxidation of metabolic substrates promotes the reduction of redox coenzymes (NAD+ and
FAD) that feed the electrons (e-) to the complexes of the respiratory chain (CI to CIV) placed in
the inner mitochondrial membrane (I.M.). Electron flow down the respiratory chain to molecular
oxygen generates the water of respiration. Electron transfer is coupled with proton pumping from
the matrix interior to the intermembrane space at CI, CIII, and CIV to generate the proton electrochemical gradient (⌬␮H+ ) across the inner membrane. ⌬␮H+ is the driving force used by the
H+ -ATP synthase (CV) for the synthesis of ATP. However, at complexes I and III, some of the electrons can combine with molecular oxygen to generate the toxic superoxide radical (dashed lines).
Q, ubiquinone; Cyt c, cytochrome c; O.M., outer membrane
biosynthesis of adenosine triphosphate (ATP) from adenosine diphosphate (ADP)
and inorganic phosphate (Pi) in the process known as oxidative phosphorylation
(OXPHOS) (Fig. 2). Whereas the metabolism of nutrient intermediates and
OXPHOS were thought to be the primary roles of mitochondria for many
decades, recent findings indicate that mitochondria play additional essential cellular
functions. The protist Giardia lamblia carries remnant organelles resembling mitochondria that, though they are unable to provide ATP, instead synthesize ironsulfur (Fe-S) clusters [7]. Fe-S clusters are essential parts of specific mitochondrial
enzymes, namely aconitase in the Krebs cycle as well as complexes of the respiratory chain [8]. Briefly, Fe-S clusters are known to carry non-heme iron, which can
both donate and accept electrons [9]. Moreover, mitochondria are central regulators and executors of cell death [10], and the acquisition of a resistant phenotype
to apoptosis is a critical aspect in the biology of tumors and its therapy [11, 12].
It is now firmly established that upon the activation of cell death by various types
of stimuli (genetic damage, ischemia, chemical and/or metabolic stresses), mitochondria release proteins (cyt c, AIF, Endo G, Smac/DIABLO, Omi/HtrA2) and/or
other molecules such as the superoxide anion (O2 . ) that participate in signaling and
Oxidative Phosphorylation and Cancer: The Ongoing Warburg Hypothesis
5
executing cell demise. It should be noted that in most cells, reactive oxygen species
(ROS) are produced by single electron transfer reactions to molecular oxygen during mitochondrial respiration (Fig. 2). In summary, mitochondria are responsible
for the production of Fe-S clusters, the conversion of primary energy substrates into
each other, the generation of large amounts of ATP, and the signaling and execution
of cell death. The major advantage of oxidative metabolism is the comparably high
energy yield, and the major disadvantage is its dependency on the availability of
oxygen and the subsequent production of ROS that could damage essential cellular
constituents.
Unaware of the above-mentioned facts, the physiologist Otto Warburg (Fig. 3
and Fig. 4) proposed as early as 1924 that malignant growth might be caused
by increased glycolysis accompanied by impaired oxygen consumption [13–16].
This hypothesis was based on his original observations that cancer tissues show
a low respiratory rate compared with that of adjacent healthy (i.e., nonmalignant)
tissues. After becoming the Nobel laureate for medicine and physiology in 1931
[17], his hypothesis on initiation of cancer obtained increasing attention and caused
widespread debate [18–22]. Warburg died in 1970 while his hypothesis was still
Fig. 3 Otto Warburg (date
unknown). (Photograph
courtesy of the Archives of
the Max Planck Society,
Berlin, Germany.)
6
M. Ristow, J.M. Cuezva
Fig. 4 Otto Warburg in his
laboratory (date unknown).
(Photograph courtesy of the
Archives of the Max Planck
Society, Berlin, Germany.)
unconfirmed, denied, and/or neglected [23, 24] mostly because of the abandonment of metabolic studies in favor of the development of molecular biology and
the discovery of oncogenes, tumor suppressor genes, and other advances in cancer
research. In any case, and after both OXPHOS and respiration had been attributed to
mitochondria, electron microscopy–based techniques were developed to study the
morphology of these organelles in tumor tissues. In summary, mitochondria from
rapidly growing cancers show less cristae and are smaller than those from betterdifferentiated cancers and normal tissues [25, 26]. Besides these structural differences in the mitochondria of the cancer cell, it has been further emphasized that
some cancer tissues also have a significant reduction in the number of mitochondria
per cell [25–27].
The abnormal high-aerobic glycolysis of tumors discovered by Warburg was confirmed by most researchers studying cancer cells, and a significant number of potential mechanisms have been put forward to explain the metabolic changes specific
to cancer cells. These can be separated into mechanisms that impinge on cellular glucose uptake and consumption by glycolysis (Fig. 1) and those affecting the
mitochondrial metabolism of pyruvate (Fig. 1 and Fig. 2). When comparing energy
metabolism of tumors and healthy tissue, profound changes have been observed in
virtually all cases of cancer or tissue types. Increased glycolysis has been found in
cancer cells at the genomic [28], proteomic [26, 29, 30–32], and metabolic levels:
Oxidative Phosphorylation and Cancer: The Ongoing Warburg Hypothesis
7
increased glucose uptake [33, 34], as well as increased production of lactate as the
end product of glycolysis [35, 36].
A modern noninvasive imaging technique called positron emission tomography
(PET) is nowadays in clinical use for diagnosing and staging cancer patients [37]
because it is remarkably sensitive to detect tumor tissue and its metastasis in situ
based on the principles of the Warburg hypothesis [32]. By injecting the nonmetabolizable glucose analogue [18 F]fluoro-2-deoxy-d-glucose (FDG) as a tracer substance
into the human body, tissues with high FDG uptake (high glucose avidity) can easily
be identified from the surrounding normal tissue in images of PET scans [38]. Using
this approach, we have recently showed that lung tumors with high glucose avidity have a profound alteration of the bioenergetic proteome [32]. The bioenergetic
proteome (BEC index) provides a signature of the tumor that informs of the preponderance of the metabolic pathways relevant for cellular energy provision at the
protein level [2, 26]. Remarkably, the analysis of the bioenergetic signature in tumor
specimens and in cells in culture [32] revealed that the downregulation of the bioenergetic function of mitochondria (Fig. 1 and Fig. 2), specifically of oxidative phosphorylation, underlies the high rates of glucose uptake and consumption by aerobic
glycolysis in the carcinomas (Fig. 1) [32]. These results have provided the first indications that integrate molecular and functional data to confirm Warburg’s initial
observations [32].
In a recent review, Gatenby and Gillies put forward an evolutionary model to
explain the apparently high glycolytic rates of cancer cells [39]. Under conditions of
extensive hypoxia, which most likely occurs during hyperplastic growth, evolutionary selection favors cells with elevated glycolytic capacity [40]. Moreover, increased
glycolysis causes increased lactate production and subsequent lactate-dependent
acidification. Selection is in favor of those cells that exhibit increased resistance
toward an unfavorable acidic environment. Therefore, a premalignant cell will be
less dependent on oxygen and is still able to expand under conditions that would
kill a nonmalignant cell. In support of Warburg’s hypothesis, this simple model provides an explanation on the benefit of cancer cells from an energetic trade-off (i.e.,
preference for nonmitochondrial energy supply).
Numerous experimental findings further emphasize this assumption. Induction of
glucose-specific transporters (GLUTs), and, most importantly GLUT1, was several
times correlated with tumor aggressiveness [41–43]. Moreover, additional evidence
links a number of established cancer genes to control of glycolytic flux. In this
regard, it has been shown that the c-myc oncogene induces expression of GLUT1
and a number of additional genes involved in glycolytic flux, including glucose-6phosphate isomerase, phosphofructokinase, glyceraldehyde-3-phosphate dehydrogenase, phosphoglycerate kinase, lactate dehydrogenase, and others (Fig. 1) [44].
Of interest, c-Myc also acts to transactivate the lactate dehydrogenase A (LDHA)
gene which is necessary for c-myc–mediated transformation in fibroblasts [35].
Accordingly, expression of other oncogenes (ras, src, PDGF) causes increased glucose uptake in malignant cells [34, 35, 45]. Additionally, and acting as c-myc itself,
such cells were capable of coactivating LDHA expression via proposed binding to
a carbohydrate responsive element found in the LDHA promoter region [46].
8
M. Ristow, J.M. Cuezva
Another relevant mechanism to regulate glycolysis and metabolic adaptation to
oxygen depletion is that mediated by the hypoxia inducible factor 1 alpha (HIF-1␣).
Subsequent to its identification in 1992 [47], a significant body of evidence has
emerged proposing that HIF-1␣ might be relevant for many of the effects as suggested in the initial Warburg hypothesis [48]. HIF-1␣ forms dimers with its partner HIF-1␤ to become an active transcription factor binding to promoter elements
of several genes including those encoding for proteins of glucose transport [49]
and activating the expression of several enzymes of the glycolytic pathway (Fig. 1)
including LDHA [50, 51]. Furthermore, the active HIF-dimer promotes angiogenesis and erythropoiesis [47, 52]. HIF-1␣ is constitutively expressed in most tissues
and degraded under conditions of normal oxygen pressure. In contrast, when oxygen
partial pressure is decreased, degradation of HIF-1␣ is blocked causing increased
transcriptional activity to regulate the respective target genes following changes
in oxygen availability [53]. Accordingly, deregulated HIF-1␣ or sustained stability under conditions of normal oxygen supply has been linked to numerous types of
cancer in past years [54]. Furthermore, the ubiquitin ligase responsible for HIF-1␣
inactivation, the von Hippel–Lindau tumor suppressor (VHL) protein, has been affiliated with the von Hippel–Lindau syndrome, which promotes neuronal and kidney
cancers [55]. Moreover, downregulation of pyruvate dehydrogenase might lead to
accumulation of pyruvate and increased lactate production, which contributes to stabilization of HIF-1␣ [56]. Hence, although hypoxia might play a key role in tumor
development, it seems not to be unambiguously required for sustained increases
in glycolysis. Putatively, this may promote a vicious cycle of enhanced glycolysis
leading to increased accumulation of pyruvate, which subsequently may stimulate
HIF-1␣ and lastly further increase the expression of glycolytic enzymes [57].
Another candidate pathway controlling the glycolysis-dependent phenotype
of malignancies, the Akt/PBK (protein kinase B) pathway, has been proposed.
Akt/PKB has long been known for its capability of inducing glucose incorporation and glycolytic flux independently of malignancy-promoting capacity [58]. In
malignant cells, elevated glycolytic flux was linked to activation of Akt/PKB [59].
Accordingly, blastoma cells lacking constitutive activation of Akt/PKB did not display altered glycolytic flux, whereas expression of Akt/PBK promoted cell death
under conditions of glucose deprivation. Moreover, it has recently been shown that
the overexpression of Akt/PBK in noninvasive radial-growth melanoma cells triggers an increase in protein markers of glycolysis and in the glycolytic flux of the
melanoma concurrently with its transformation to the aggressive vertical-growth
invasive behavior [60]. Altogether, these results suggest the possibility that deregulated Akt/PKB signaling might contribute to the Warburg effect in cancer cells.
Additionally, interactions of glycolytic enzymes with the well-characterized p53
tumor suppressor protein have been described, further promoting the relevance of
cancer-related genes in causing a potential metabolic shift toward increased glycolytic rate. In this regard, a putative role for hexokinase has been proposed [61]. In
hepatoma cells, hexokinase was closely associated with the mitochondria, which
was not observed in normal (i.e., untransformed) liver cells [62]. A subsequent
study revealed that mutations in the p53 tumor suppressor protein might activate
Oxidative Phosphorylation and Cancer: The Ongoing Warburg Hypothesis
9
hexokinase II expression via binding to specific sites of the corresponding hexokinase promoter [63, 64]. In further support of these interesting observations, a protein
named TP53-induced glycolysis and apoptosis regulator (TIGAR ), an isoform of
phosphofructokinase that can be induced by p53, was identified in recent years
[65]. TIGAR acts as an inhibitor of glycolysis by lowering levels of fructose-2,6bisphosphate. Accordingly, in cases of p53 mutations as found in many cancers,
activity of TIGAR is downregulated promoting an increase in glycolysis. In summary, a number of cancer-related genes have been shown to promote alterations on
the glycolytic phenotype that could explain that particular aspect of the Warburg
effect.
Aspects other than deregulated glycolysis underscore that oxidative metabolism
itself is also impaired in malignant cells, thereby representing another key aspect
of Warburg’s hypothesis. The phenomenon of impaired mitochondrial activity [13,
15, 25] appears to be a key feature of numerous cancer specimens [26, 29, 31, 32,
66, 67]. Accordingly, mutations of mitochondrial DNA have repeatedly been associated with induction of tumor growth on a hypothetical basis [67, 68] and later
on a factual basis [69–73]. Complexes of the mitochondrial respiratory chain (Fig.
2) including cytochrome c oxidase (COX) exhibit diminished activity in cancers,
mainly due to decreased expression of the subunits encoded by the mitochondrial
DNA [74]. Likewise, mutations in nuclear genes that affect the bioenergetic function of mitochondria, and thus compromise the provision of metabolic energy, have
also been shown to predispose to different types of inherited neoplasia syndromes
[75–77].
Moreover, over the past years it has been shown that most prevalent human
tumors fit the original Warburg hypothesis because they show a decrease in the
expression of ␤-F1-ATPase, which is the catalytic subunit of the mitochondrial H+ ATP synthase and thus a bottleneck of oxidative phosphorylation (CV in Fig. 2),
concurrently with an increased expression of several makers of the glycolytic pathway (26, 29, 31 and references therein). This proteomic feature, which was defined
as “the bioenergetic signature” of cancer [26], illustrates both alterations in the
bioenergetic proteome of the mitochondrion and on the overall mitochondrial activity of the cancer cell [2] and provides markers of the two mechanisms that could
interfere with mitochondrial activity in the various types of cancer cells [2]. This
is especially relevant because depending upon the tissue type considered, a general
reduction in mitochondrial mass or a specific reduction on the expression of the ATP
synthase is associated with progression of the neoplasia [2, 26]. Interestingly, these
recent findings perfectly complement earlier findings on reduced ATP(synth)ase
activity in rodent models of liver carcinogenesis [78]. Additionally, Racker and coworkers also emphasized the putative relevance of ATP(synth)ase in the control of
tumor proliferation [79, 80].
As indicated previously, we have recently documented both in vivo using FDGPET imaging in lung cancer patients and in vitro that the rates of glucose capture by lung tumors and of glucose utilization by aerobic glycolysis in cancer
cells depend on the expression level of the ␤-F1-ATPase protein and on the activity of oxidative phosphorylation [32]. These results provide strong support to the
10
M. Ristow, J.M. Cuezva
original Warburg hypothesis and to the principle of metabolic regulation known as
the Pasteur effect [81]. This principle of metabolism, which basically states that “the
metabolic flux of glycolysis in eukaryotic cells depends on the energy provided in
the form of ATP by mitochondrial activity,” is what was used by Warburg as the
“first stone” for the building of his seminal and outstanding hypothesis in the early
dates of biochemistry [2]. Remarkably, the bioenergetic signature has been shown
to provide a relevant marker of disease progression in colon [26], lung [30, 32], and
breast [31] cancer patients. Furthermore, it has been documented that the bioenergetic signature also provides a marker of the cellular response to chemotherapy in
colon [82] and liver [83] cancer cells that might allow its further use as a predictive
marker of therapeutic intervention and/or as a target of future cancer treatments.
Because it appears that alterations in mitochondrial energetic physiology contribute to the Warburg effect, that is, an impaired oxidative phosphorylation in mitochondria promotes the upregulation of glycolysis (Fig. 1), at least two questions
arise: (i) what are the mechanism(s) that impair the OXPHOS activity of mitochondria, and (ii) what is the advantage and mechanistic contribution of the downregulation of OXPHOS for the cancer cell? These will be partially addressed in the
following.
Supported by previous findings on p53 as mentioned before, it has been shown
that this tumor suppressor protein may directly be involved in regulating mitochondrial oxygen consumption through regulation of COX activity [84]. Accordingly,
cancer cell lines lacking p53 showed decreased rates of OXPHOS as reflected by
impaired ATP synthesis and decreased respiratory rate, and maintenance of cellular ATP levels was dependent on a concurrent increase in glycolysis [85, 86]. This
specific regulation was promoted by Synthesis of Cytochrome c Oxidase 2 SCO2,
which has previously been shown to be involved in the assembly of complex IV of
the respiratory chain [86]. Nevertheless, it remains to be shown whether this particular activation can be directly linked to mitochondrial activity. Therefore, this
tumor suppressor, p53, not only regulates glycolysis as shown above but possibly
also directly affects mitochondrial oxygen consumption.
Not only p53 has been shown to influence both glycolysis and mitochondrial activity. Similar conjunct activities have been put forward for both HIF-1␣ and Akt/PKB
signaling pathways. In this regard, HIF-1␣ has previously been demonstrated to activate pyruvate dehydrogenase kinase (PDK), which is known to inactivate pyruvate
dehydrogenase, preventing the decarboxylation of pyruvate and hence its further oxidation in the mitochondria having a negative impact on mitochondrial oxygen consumption [87, 88]. Likewise, the above-mentioned tumor suppressor protein VHL
not only ubiquitinates HIF-1␣ but also promotes mitochondrial biogenesis and activity. Accordingly, loss of VHL, as repeatedly described in various types of cancer,
might possibly impair mitochondrial activity strongly promoting the Warburg effect
[66]. Moreover, reduced OXPHOS capacity has previously been suggested to promote phosphorylation and hence activation of Akt [89], a well-defined oncogene, in
a Phosphatase and Tensin Homolog PTEN-dependent manner [90].
An additional alternative to explain the “Warburg phenotype” of the cancer cell
is the one we have recently put forward [32] after the characterization of the biogenesis and functional development of mitochondria during the cell cycle [91]. We have
Oxidative Phosphorylation and Cancer: The Ongoing Warburg Hypothesis
11
suggested that the downregulation of oxidative phosphorylation and the concurrent
upregulation of glycolysis is the phenotype that identifies the energetic metabolism
of most proliferating cells. In other words, glycolysis is the metabolic pathway that
supports proliferation and tumor growth. In this regard, it is well established that
lymphocytes switch to glycolysis as the energy-producing pathway during proliferation [92], and progression through the S/G2/M phases of the cell cycle, the socalled reductive phases of the “metabolic cycle” in which most of the mitochondrial
components are being synthesized [91], is supported by nonrespiratory modes of
energy generation [93]. In agreement with these findings, it has recently been shown
that cyclin D1, which is involved in the phosphorylation and inactivation of the
retinoblastoma protein marking in this way the entry of cells into the S phase,
represses mitochondrial function in vivo [94, 95]. We presume that the effects that
oncogenes and tumor suppressors could have on mitochondrial activity might be
superimposed to the natural metabolic phenotype of proliferating cells. Mechanistically, and by analogy with findings on the biogenesis of mitochondria in rat hepatomas [27], in fetal rat liver [96–98] and during the cell cycle [91], the repression
of ␤-F1-ATPase expression in human tumors [26, 29–31] could be exerted by translation masking (silencing) of the ␤-F1-ATPase mRNA, a condition that might also
affect other OXPHOS mRNAs of the cancer cell [2].
According to Warburg’s hypothesis, impaired mitochondrial function should be
considered a key property of cancer cells (i.e., an additional hallmark of cancer to be
added to the list) [99]. Strongly supporting this hypothesis, it was shown that induction of mitochondrial energy conversion by forced expression of the mitochondrial
protein frataxin [100] leads to reduced growth of cancer cells [101]. Frataxin has
been shown to be involved in Fe-S cluster biogenesis [102] and induces OXPHOS
[100]. In this regard, and as a novel pathway connecting induction of mitochondrial
metabolism and decreased growth rates, mitogen activated protein kinase p38 has
been proposed [101, 103]. This protein p38 has repeatedly been observed to act as
an efficient tumor suppressor protein [104, 105]. In full accordance with these observations, hepatocyte-specific inactivation of murine frataxin reduces the activity of
Fe-S cluster–containing enzymes within the mitochondria, promoting hepatic adenoma formation in such mice [103]. Consistently, defective Krebs cycle activity has
been observed in colorectal cancer samples including reduced activity of the Fe-S
cluster–containing enzyme aconitase [106]. Taken together, these findings indicate
that activation of mitochondrial energy metabolism has a tumor suppressor effect that
opposes the Warburg phenotype and could be a therapeutic tool in cancer therapy.
Indeed, the execution of cell death requires an efficient oxidative phosphorylation [107–110], the molecular components of the H+ -ATP synthase being specifically required [111–113]. In this regard, we have recently shown that the activity of the H+ -ATP synthase is necessary for the efficient execution of apoptosis
in cells that depend on oxidative phosphorylation for the provision of metabolic
energy, whereas it is dispensable in those cells that rely heavily on glycolysis [83].
The role of the H+ -ATP synthase in the execution of cell death is mediated by
controlling the generation of ROS [83], which in turn promote a severe oxidative damage on mitochondrial proteins, favoring in this way the release of apoptogenic molecules from the organelle [83]. Consistently, highly glycolytic cells
12
M. Ristow, J.M. Cuezva
with an arrested mitochondrial biogenesis and thus with no dependence on oxidative phosphorylation for energy provision do not produce ROS and are resistant to
mitochondria-geared cell death stimuli [83]. Altogether, these results provide additional evidence linking metabolism to cell death [114–117] and support strongly
that the acquisition of a Warburg phenotype is another strategy of the cancer cell in
order to ensure its perpetuation.
In recent years, AMP activated protein kinase (AMPK) has been established as
a key regulator of cellular energy balance controlling a significant number of pathways that control cellular ATP content [118]. AMPK is activated under conditions of
low energy supply, induced by stressors like reduced nutrient availability or hypoxia
[118]. Activation of AMPK is initiated by the upstream tumor suppressor kinase,
LKB1, which phosphorylates AMPK [119]. Of note, a mutation of LKB1 has been
shown to cause Peutz-Jeghers syndrome, a disease that has been linked to increased
incidence of cancers. Moreover, two other downstream targets of AMPK, the tumor
suppressors TSC1 and 2, have been shown to be mutated in tuberous sclerosis [120].
One of the major outcomes of AMPK and/or activation of its downstream substrates
is the induction of mitochondrial biogenesis and/or activity [121]. Not surprisingly,
AMPK appears to be relevant in the control of protein synthesis and proliferation
[122], and may activate p53 causing cell cycle arrest and/or cellular senescence in
states of decreased glucose availability [123]. Given the connection of AMPK on
the one hand and several tumor suppressors, including TSCs and LKB1 (see above),
on the other hand, as well as the fact that AMPK itself is of eminent relevance to
mitochondrial activity and energy balance, disruption of this pathway might be considered of importance in regard to the molecular causes of the Warburg effect in
states of cancer and malignant growth.
In summary, though the Warburg hypothesis was formulated more than 80 years
ago, nowadays it still (or rather again) is a subject of extensive debate. The recent
findings on the metabolic signature of tumors and malignant cells on the one hand
and several molecular pathways linked to derangements of cellular energy flux in
transformed state on the other hand strongly support the overdue importance of one
of the most long-standing hypotheses in oncology—Warburg’s hypothesis.
Acknowledgments We thank members of our laboratories for their contributions and encouraging
discussions on the subject of this review. This review article was written while the research activity
in the authors’ laboratory was supported by grants from the German government to M.R. and
from the Ministerio de Educaci´on y Ciencia (BFU2007-65253) and Fundaci´on Mutua Madrile˜na
to J.M.C.
References
1. Ristow M. Oxidative metabolism in cancer growth. Curr Opin Clin Nutr Metab Care 2006;
9:339–345.
2. Cuezva JM, Sanchez-Arago M, Sala S, Blanco-Rivero A, Ortega AD. A message emerging
from development: the repression of mitochondrial beta-F1-ATPase expression in cancer.
J Bioenerg Biomembr 2007; 39:259–265.
Oxidative Phosphorylation and Cancer: The Ongoing Warburg Hypothesis
13
3. Lichtenstein AH, Appel LJ, Brands M, et al. Diet and lifestyle recommendations revision
2006: a scientific statement from the American Heart Association Nutrition Committee. Circulation 2006; 114:82–96.
4. Pasteur L. Influence de l oxygene sur le developpement de la levure et la fermentation
alcoolique. Bulletin de la Societe Chimique de Paris 1861; 79–80.
5. Pfeiffer T, Schuster S, Bonhoeffer S. Cooperation and competition in the evolution of ATPproducing pathways. Science 2001; 292:504–507.
6. Boyer PD. The ATP synthase. A splendid molecular machine. Annu Rev Biochem 1997;
66:717–749.
7. Tovar J, Leon-Avila G, Sanchez LB, et al. Mitochondrial remnant organelles of Giardia
function in iron-sulphur protein maturation. Nature 2003; 426:172–176.
8. Lill R, M¨uhlenhoff U. Iron-sulfur-protein biogenesis in eukaryotes. Trends Biochem Sci
2005; 30:133–141.
9. Rouault TA, Klausner RD. Iron-sulfur clusters as biosensors of oxidants and iron. Trends
Biochem Sci 1996; 21:174–177.
10. Wang X. The expanding role of mitochondria in apoptosis. Genes Dev 2001; 15:
2922–2933.
11. Green DR, Reed JC. Mitochondria and apoptosis. Science 1998; 281:1309–1312.
12. Jaattela M. Multiple cell death pathways as regulators of tumour initiation and progression.
Oncogene 2004; 23:2746–2756.
13. Warburg O. On the origin of cancer cells. Science 1956; 123:309–314.
14. Warburg O. The metabolism of tumours. London: Constable, 1930.
15. Warburg O. On respiratory impairment in cancer cells. Science 1956; 124:269–270.
¨
16. Warburg O, Posener K, Negelein E. Uber
den Stoffwechsel der Tumoren [On metabolism of
tumors]. Biochemische Zeitschrift 1924; 152:319–344.
17. Anonymous. The Nobel prize in physiology or medicine 1931. Available at:
http://nobelprizeorg/medicine/laureates/1931/.
18. Burk D. A colloquial consideration on the Pasteur and Neo-Pasteur effects. Cold Spring
Harbor Symposia on Quantitative Biology 1939; 7:420–460.
19. Burk D, Schade AL. On respiratory impairment in cancer cells. Science 1956; 124:
270–272.
20. Burk D, Woods M. Newer aspects of glucose fermentation in cancer growth and control.
Arch Geschwulstforsch 1967; 28:305–319.
21. Weinhouse S. On respiratory impairment in cancer cells. Science 1956; 124:267–269.
22. Zu XL, Guppy M. Cancer metabolism: facts, fantasy, and fiction. Biochem Biophys Res
Commun 2004; 313:459–465.
23. Warburg O. Annual meeting of Nobelists at Lindau, Germany, 1966. English edition by
D. Burk. Bethesda, MD: National Cancer Institute. Available at: http://www.hopeforcancer.
com/OxyPlus.htm.
24. Krebs H. Otto Warburg: cell physiologist, biochemist and eccentric. Oxford, UK: Clarendon,
1981.
25. Pedersen PL. Tumor mitochondria and the bioenergetics of cancer cells. Prog Exp Tumor
Res 1978; 22:190–274.
26. Cuezva JM, Krajewska M, de Heredia ML, et al. The bioenergetic signature of cancer: a
marker of tumor progression. Cancer Res 2002; 62:6674–6681.
27. L´opez de Heredia M, Izquierdo JM, Cuezva JM. A conserved mechanism for controlling the
translation of beta-F1-ATPase mRNA between the fetal liver and cancer cells. J Biol Chem
2000; 275:7430–7437.
28. Altenberg B, Greulich KO. Genes of glycolysis are ubiquitously overexpressed in 24 cancer
classes. Genomics 2004; 84:1014–1020.
29. Isidoro A, Martinez M, Fernandez PL, et al. Alteration of the bioenergetic phenotype of
mitochondria is a hallmark of breast, gastric, lung and oesophageal cancer. Biochem J 2004;
378:17–20.
14
M. Ristow, J.M. Cuezva
30. Cuezva JM, Chen G, Alonso AM, et al. The bioenergetic signature of lung adenocarcinomas is a molecular marker of cancer diagnosis and prognosis. Carcinogenesis 2004; 25:
1157–1163.
31. Isidoro A, Casado E, Redondo A, et al. Breast carcinomas fulfill the Warburg hypothesis and
provide metabolic markers of cancer prognosis. Carcinogenesis 2005; 26:2095–2104.
32. L´opez-R´ıos F, S´anchez-Arag´o M, Garc´ıa-Garc´ıa E, et al. Loss of the mitochondrial bioenergetic capacity underlies the glucosa avidity of carcinomas. Cancer Res 2007; 67:9013–9017.
33. Birnbaum MJ, Haspel HC, Rosen OM. Transformation of rat fibroblasts by FSV rapidly
increases glucose transporter gene transcription. Science 1987; 235:1495–1498.
34. Flier JS, Mueckler MM, Usher P, Lodish HF. Elevated levels of glucose transport and transporter messenger RNA are induced by ras or src oncogenes. Science 1987; 235:1492–1495.
35. Shim H, Dolde C, Lewis BC, et al. c-Myc transactivation of LDH-A: implications for tumor
metabolism and growth. Proc Natl Acad Sci USA 1997; 94:6658–6663.
36. Ziegler A, von Kienlin M, Decorps M, Remy C. High glycolytic activity in rat glioma
demonstrated in vivo by correlation peak 1H magnetic resonance imaging. Cancer Res 2001;
61:5595–5600.
37. Rigo P, Paulus P, Kaschten BJ, et al. Oncological applications of positron emission tomography with fluorine-18 fluorodeoxyglucose. Eur J Nucl Med 1996; 23:1641–1674.
38. Gambhir SS. Molecular imaging of cancer with positron emission tomography. Nat Rev
Cancer 2002; 2:683–693.
39. Gatenby RA, Gillies RJ. Why do cancers have high aerobic glycolysis? Nat Rev Cancer
2004; 4:891–899.
40. Raghunand N, Gatenby RA, Gillies RJ. Microenvironmental and cellular consequences of
altered blood flow in tumours. Br J Radiol 2003; 76(Spec No 1):S11–22.
41. Grover-McKay M, Walsh SA, Seftor EA, Thomas PA, Hendrix MJ. Role for glucose transporter 1 protein in human breast cancer. Pathol Oncol Res 1998; 4:115–120.
42. Sakashita M, Aoyama N, Minami R, et al. Glut1 expression in T1 and T2 stage colorectal
carcinomas: its relationship to clinicopathological features. Eur J Cancer 2001; 37:204–209.
43. Younes M, Lechago LV, Lechago J. Overexpression of the human erythrocyte glucose transporter occurs as a late event in human colorectal carcinogenesis and is associated with an
increased incidence of lymph node metastases. Clin Cancer Res 1996; 2:1151–1154.
44. Osthus RC, Shim H, Kim S, et al. Deregulation of glucose transporter 1 and glycolytic gene
expression by c-Myc. J Biol Chem 2000; 275:21797–21800.
45. Govindarajan B, Shah A, Cohen C, et al. Malignant transformation of human cells by
constitutive expression of platelet-derived growth factor-BB. J Biol Chem 2005; 280:
13936–13943.
46. Dang CV, Lewis BC, Dolde C, Dang G, Shim H. Oncogenes in tumor metabolism, tumorigenesis, and apoptosis. J Bioenerg Biomembr 1997; 29:345–354.
47. Semenza GL, Wang GL. A nuclear factor induced by hypoxia via de novo protein synthesis binds to the human erythropoietin gene enhancer at a site required for transcriptional
activation. Mol Cell Biol 1992; 12:5447–5454.
48. Robey IF, Lien AD, Welsh SJ, Baggett BK, Gillies RJ. Hypoxia-inducible factor-1alpha and
the glycolytic phenotype in tumors. Neoplasia 2005; 7:324–330.
49. Ebert BL, Firth JD, Ratcliffe PJ. Hypoxia and mitochondrial inhibitors regulate expression of glucose transporter-1 via distinct cis-acting sequences. J Biol Chem 1995; 270:
29083–29089.
50. Firth JD, Ebert BL, Ratcliffe PJ. Hypoxic regulation of lactate dehydrogenase A. Interaction between hypoxia-inducible factor 1 and cAMP response elements. J Biol Chem 1995;
270:21021–21027.
51. Semenza GL, Roth PH, Fang HM, Wang GL. Transcriptional regulation of genes encoding
glycolytic enzymes by hypoxia-inducible factor 1. J Biol Chem 1994; 269:23757–23763.
52. Maxwell PH, Dachs GU, Gleadle JM, et al. Hypoxia-inducible factor-1 modulates gene
expression in solid tumors and influences both angiogenesis and tumor growth. Proc Natl
Acad Sci USA 1997; 94:8104–8109.
Oxidative Phosphorylation and Cancer: The Ongoing Warburg Hypothesis
15
53. Semenza GL. HIF-1, O(2), and the 3 PHDs: how animal cells signal hypoxia to the nucleus.
Cell 2001; 107:1–3.
54. Semenza GL. HIF-1 mediates the Warburg effect in clear cell renal carcinoma. J Bioenerg
Biomembr 2007; 39:231–234.
55. McKusick VA, Kniffin CL, Tiller GE, et al. Online Mendelian Inheritance in Man:
von Hippel-Lindau syndrome (OMIM 193300). Available at: http://www.ncbi.nlm.nih.gov/
entrez/dispomim.cgi?id=193300.>
56. Koukourakis MI, Giatromanolaki A, Sivridis E, Gatter KC, Harris AL. Pyruvate dehydrogenase and pyruvate dehydrogenase kinase expression in non small cell lung cancer and
tumor-associated stroma. Neoplasia 2005; 7:1–6.
57. Lu H, Dalgard CL, Mohyeldin A, McFate T, Tait AS, Verma A. Reversible inactivation of
HIF-1 prolyl hydroxylases allows cell metabolism to control basal HIF-1. J Biol Chem 2005;
280:41928–41939.
58. Rathmell JC, Fox CJ, Plas DR, Hammerman PS, Cinalli RM, Thompson CB. Akt-directed
glucose metabolism can prevent Bax conformation change and promote growth factorindependent survival. Mol Cell Biol 2003; 23:7315–7328.
59. Elstrom RL, Bauer DE, Buzzai M, et al. Akt stimulates aerobic glycolysis in cancer cells.
Cancer Res 2004; 64:3892–3899.
60. Govindarajan B, Sligh JE, Vincent BJ, et al. Overexpression of Akt converts radial growth
melanoma to vertical growth melanoma. J Clin Invest 2007; 117:719–729.
61. Pedersen PL. Warburg, me and hexokinase 2: multiple discoveries of key molecular events
underlying one of cancers most common phenotypes, the ”Warburg Effect,” i.e., elevated
glycolysis in the presence of oxygen. J Bioenerg Biomembr 2007; 39:211–222.
62. Bustamante E, Morris HP, Pedersen PL. Energy metabolism of tumor cells. Requirement
for a form of hexokinase with a propensity for mitochondrial binding. J Biol Chem 1981;
256:8699–8704.
63. Mathupala SP, Heese C, Pedersen PL. Glucose catabolism in cancer cells. The type II hexokinase promoter contains functionally active response elements for the tumor suppressor
p53. J Biol Chem 1997; 272:22776–22780.
64. Smith TA, Sharma RI, Thompson AM, Paulin FE. Tumor 18F-FDG incorporation is
enhanced by attenuation of P53 function in breast cancer cells in vitro. J Nucl Med 2006;
47:1525–1530.
65. Bensaad K, Tsuruta A, Selak MA, et al. TIGAR, a p53-inducible regulator of glycolysis and
apoptosis. Cell 2006; 126:107–120.
66. Hervouet E, Demont J, Pecina P, et al. A new role for the von Hippel-Lindau tumor suppressor protein: stimulation of mitochondrial oxidative phosphorylation complex biogenesis.
Carcinogenesis 2005; 26:531–539.
67. Wu M, Neilson A, Swift AL, et al. Multiparameter metabolic analysis reveals a close link
between attenuated mitochondrial bioenergetic function and enhanced glycolysis dependency in human tumor cells. Am J Physiol Cell Physiol 2007; 292:C125–136.
68. Hoberman HD. Is there a role for mitochondrial genes in carcinogenesis? Cancer Res 1975;
35:3332–3335.
69. Baggetto LG. Role of mitochondria in carcinogenesis. Eur J Cancer 1992; 29A:156–159.
70. Carew JS, Huang P. Mitochondrial defects in cancer. Mol Cancer 2002; 1:9.
71. Petros JA, Baumann AK, Ruiz-Pesini E, et al. mtDNA mutations increase tumorigenicity in
prostate cancer. Proc Natl Acad Sci USA 2005; 102:719–724.
72. Singh KK. Mitochondrial dysfunction is a common phenotype in aging and cancer. Ann N
Y Acad Sci 2004; 1019:260–264.
73. Taylor RW, Turnbull DM. Mitochondrial DNA mutations in human disease. Nat Rev Genet
2005; 6:389–402.
74. Herrmann PC, Herrmann EC. Oxygen metabolism and a potential role for cytochrome c
oxidase in the Warburg effect. J Bioenerg Biomembr 2007; 39:247–250.
75. Baysal BE, Ferrell RE, Willett-Brozick JE, et al. Mutations in SDHD, a mitochondrial complex II gene, in hereditary paraganglioma. Science 2000; 287:848–851.
16
M. Ristow, J.M. Cuezva
76. Habano W, Sugai T, Nakamura S, Uesugi N, Higuchi T, Terashima M, Horiuchi S. Reduced
expression and loss of heterozygosity of the SDHD gene in colorectal and gastric cancer.
Oncol Rep 2003; 10:1375–1380.
77. Neumann HP, Pawlu C, Peczkowska M, et al. Distinct clinical features of paraganglioma
syndromes associated with SDHB and SDHD gene mutations. JAMA 2004; 292:943–951.
78. Deml E, Oesterle D, Wolff T, Greim H. Age-, sex-, and strain-dependent differences in the
induction of enzyme-altered islands in rat liver by diethylnitrosamine. J Cancer Res Clin
Oncol 1981; 100:125–134.
79. Racker E. Bioenergetics and the problem of tumor growth. Am Sci 1972; 60:56–63.
80. Racker E, Spector M. Warburg effect revisited: merger of biochemistry and molecular biology. Science 1981; 213:303–307.
81. Lehninger AL. Biochemistry. New York: Worth Publishers, 1970.
82. Shin YK, Yoo BC, Chang HJ, et al. Down-regulation of mitochondrial F1F0-ATP synthase
in human colon cancer cells with induced 5-fluorouracil resistance. Cancer Res 2005; 65:
3162–3170.
83. Santamaria G, Martinez-Diez M, Fabregat I, Cuezva JM. Efficient execution of cell death in
non-glycolytic cells requires the generation of ROS controlled by the activity of mitochondrial H+-ATP synthase. Carcinogenesis 2006; 27:925–935.
84. Zhou S, Kachhap S, Singh KK. Mitochondrial impairment in p53-deficient human cancer
cells. Mutagenesis 2003; 18:287–292.
85. Ma W, Sung HJ, Park JY, Matoba S, Hwang PM. A pivotal role for p53: balancing aerobic
respiration and glycolysis. J Bioenerg Biomembr 2007; 39:243–246.
86. Matoba S, Kang JG, Patino WD, et al. p53 regulates mitochondrial respiration. Science
2006; 312:1650–1653.
87. Kim JW, Tchernyshyov I, Semenza GL, Dang CV. HIF-1-mediated expression of pyruvate
dehydrogenase kinase: a metabolic switch required for cellular adaptation to hypoxia. Cell
Metab 2006; 3:177–185.
88. Papandreou I, Cairns RA, Fontana L, Lim AL, Denko NC. HIF-1 mediates adaptation to
hypoxia by actively downregulating mitochondrial oxygen consumption. Cell Metab 2006;
3:187–197.
89. Pelicano H, Xu RH, Du M, et al. Mitochondrial respiration defects in cancer cells cause
activation of Akt survival pathway through a redox-mediated mechanism. J Cell Biol 2006;
175:913–923.
90. Stiles B, Groszer M, Wang S, Jiao J, Wu H. PTENless means more. Dev Biol 2004; 273:
175–184.
91. Martinez-Diez M, Santamaria G, Ortega AD, Cuezva JM. Biogenesis and dynamics of mitochondria during the cell cycle: significance of 3 UTRs. PLoS ONE 2006; 1:e107.
92. Wang T, Marquardt C, Foker J. Aerobic glycolysis during lymphocyte proliferation. Nature
1976; 261:702–705.
93. Chen Z, Odstrcil EA, Tu BP, McKnight SL. Restriction of DNA replication to the reductive
phase of the cell cycle protects genome integrity. Science 2007; 316:1916–1919.
94. Sakamaki T, Casimiro MC, Ju X, et al. Cyclin d1 determines mitochondrial function in vivo.
Mol Cell Biol 2006; 26:5449–5469.
95. Wang C, Li Z, Lu Y, et al. Cyclin D1 repression of nuclear respiratory factor 1 integrates nuclear DNA synthesis and mitochondrial function. Proc Natl Acad Sci USA 2006;
103:11567–11572.
96. Luis AM, Izquierdo JM, Ostronoff LK, Salinas M, Santar´en JF, Cuezva JM. Translational
regulation of mitochondrial differentiation in neonatal rat liver. Specific increase in the translational efficiency of the nuclear-encoded mitochondrial beta-F1-ATPase mRNA. J Biol
Chem 1993; 268:1868–1875.
97. Izquierdo JM, Ricart J, Ostronoff LK, Egea G, Cuezva JM. Changing patterns of transcriptional and post-transcriptional control of beta-F1-ATPase gene expression during mitochondrial biogenesis in liver. J Biol Chem 1995; 270:10342–10350.
Oxidative Phosphorylation and Cancer: The Ongoing Warburg Hypothesis
17
98. Izquierdo JM, Cuezva JM. Control of the translational efficiency of beta-F1-ATPase mRNA
depends on the regulation of a protein that binds the 3 untranslated region of the mRNA.
Mol Cell Biol 1997; 17:5255–5268.
99. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000; 100:57–70.
100. Ristow M, Pfister MF, Yee AJ, et al. Frataxin activates mitochondrial energy conversion and
oxidative phosphorylation. Proc Natl Acad Sci USA 2000; 97:12239–12243.
101. Schulz TJ, Thierbach R, Voigt A, et al. Induction of oxidative metabolism by mitochondrial
frataxin inhibits cancer growth: Otto Warburg revisited. J Biol Chem 2006; 281:977–981.
102. M¨uhlenhoff U, Richhardt N, Ristow M, Kispal G, Lill R. The yeast frataxin homolog Yfh1p
plays a specific role in the maturation of cellular Fe/S proteins. Hum Mol Genet 2002;
11:2025–2036.
103. Thierbach R, Schulz TJ, Isken F, et al. Targeted disruption of hepatic frataxin expression
causes impaired mitochondrial function, decreased life span, and tumor growth in mice.
Hum Mol Genet 2005; 14:3857–3864.
104. Bulavin DV, Fornace AJ Jr. p38 MAP kinase s emerging role as a tumor suppressor. Adv
Cancer Res 2004; 92:95–118.
105. Timofeev O, Lee TY, Bulavin DV. A subtle change in p38 MAPK activity is sufficient to
suppress in vivo tumorigenesis. Cell Cycle 2005; 4:118120.
106. Bi X, Lin Q, Foo TW, et al. Proteomic analysis of colorectal cancer reveals alterations in
metabolic pathways: mechanism of tumorigenesis. Mol Cell Proteomics 2006; 5:1119–1130.
107. Dey R, Moraes CT. Lack of oxidative phosphorylation and low mitochondrial membrane
potential decrease susceptibility to apoptosis and do not modulate the protective effect of
Bcl-x(L) in osteosarcoma cells. J Biol Chem 2000; 275:7087–7094.
108. Kim JY, Kim YH, Chang I, et al. Resistance of mitochondrial DNA-deficient cells to TRAIL:
role of Bax in TRAIL-induced apoptosis. Oncogene 2002; 21:3139–3148.
109. Park SY, Chang I, Kim JY, et al. Resistance of mitochondrial DNA-depleted cells
against cell death: role of mitochondrial superoxide dismutase. J Biol Chem 2004; 279:
7512–7520.
110. Tomiyama A, Serizawa S, Tachibana K, et al. Critical role for mitochondrial oxidative phosphorylation in the activation of tumor suppressors Bax and Bak. J Natl Cancer Inst 2006;
98:1462–1473.
111. Matsuyama S, Xu Q, Velours J, Reed JC. The mitochondrial F0F1-ATPase proton pump is
required for function of the proapoptotic protein Bax in yeast and mammalian cells. Mol
Cell 1998; 1:327–336.
112. Gross A, Pilcher K, Blachly-Dyson E, et al. Biochemical and genetic analysis of the mitochondrial response of yeast to BAX and BCL-X(L). Mol Cell Biol 2000; 20:3125–3136.
113. Harris MH, Vander Heiden MG, Kron SJ, Thompson CB. Role of oxidative phosphorylation
in Bax toxicity. Mol Cell Biol 2000; 20:3590–3596.
114. Plas DR, Thompson CB. Cell metabolism in the regulation of programmed cell death. Trends
Endocrinol Metab 2002; 13:75–78.
115. Danial NN, Gramm CF, Scorrano L, et al. BAD and glucokinase reside in a mitochondrial
complex that integrates glycolysis and apoptosis. Nature 2003; 424:952–956.
116. Azoulay-Zohar H, Israelson A, Abu-Hamad S, Shoshan-Barmatz V. In self-defence: hexokinase promotes voltage-dependent anion channel closure and prevents mitochondriamediated apoptotic cell death. Biochem J 2004; 377:347–355.
117. Vahsen N, Cande C, Briere JJ, et al. AIF deficiency compromises oxidative phosphorylation.
EMBO J 2004; 23:4679–4689.
118. Kahn BB, Alquier T, Carling D, Hardie DG. AMP-activated protein kinase: ancient energy
gauge provides clues to modern understanding of metabolism. Cell Metab 2005; 1:15–25.
119. Alessi DR, Sakamoto K, Bayascas JR. LKB1-dependent signaling pathways. Annu Rev
Biochem 2006; 75:137–163.
120. Inoki K, Zhu T, Guan KL. TSC2 mediates cellular energy response to control cell growth
and survival. Cell 2003; 115:577–590.
18
M. Ristow, J.M. Cuezva
121. Reznick RM, Shulman GI. The role of AMP-activated protein kinase in mitochondrial biogenesis. J Physiol 2006; 574:33–39.
122. Shaw RJ. Glucose metabolism and cancer. Curr Opin Cell Biol 2006; 18:598–608.
123. Jones RG, Plas DR, Kubek S, et al. AMP-activated protein kinase induces a p53-dependent
metabolic checkpoint. Mol Cell 2005; 18:283–293.
The Electron Transport Chain
and Carcinogenesis
Jean-Jacques Bri`ere, Paule B´enit and Pierre Rustin
Abstract Major metabolic changes that affect the balance between respiration and
glycolysis occur during carcinogenesis. It is therefore not surprising that it has long
been suggested that abnormal activity of the mitochondrial electron transfer chain
could play a role in the underlying pathophysiologic process. However, it has only
recently been demonstrated that the specific impairment of one component of the
electron transfer chain, namely complex II, can indeed be the primary event triggering carcinogenesis. Unexpectedly, rather than superoxide overproduction, the
organic acid imbalance resulting from the complex II blockade appears to be instrumental in the early step of tumorigenesis. Nevertheless, because an abnormal handling of oxygen is frequently suspected of being associated with electron transfer
dysfunction, a renewed interest is observed for the putative instability of the mitochondrial genome often reported in tumor tissues. In this chapter, we review and
discuss the evidence advocated for the implication of the electron transfer chain in
carcinogenesis.
Keywords Carcinogenesis · Mitochondria · Organic acids · Superoxides · Electron
transfer
1 Introduction
Mitochondria play an irreplaceable role in all aerobic organisms (i.e., the handling
of oxygen by the electron transfer chain for the mutual benefit of the cell host and
of these Proteobacteria-derived organelles) [1]. In addition, through a fascinating
and complex network of interactions, mitochondria have emerged as central executioners, or possibly even decision makers, for cell death under a large number
of conditions [2]. In keeping with this paradigm, a set of mitochondria-releasable
P. Rustin (B)
INSERM U676, Hˆopital Robert Debr´e, 75019 Paris, France
e-mail: rustin@rdebre.inserm.fr
S.P. Apte, R. Sarangarajan (eds.), Cellular Respiration and Carcinogenesis,
DOI 10.1007/978-1-59745-435-3 2,
C Humana Press, a part of Springer Science+Business Media, LLC 2009
19
20
J.-J. Bri´ere et al.
factors have been shown to be bifunctional proteins [3–5]. Involved in electron
transfer chain activity or maintenance, both cytochrome c and apoptosis inducing
factor (AIF) readily trigger apoptosis when released in the cytosol upon opening of
the elusive mitochondrial permeability transition pore (MPTP) [6]. Furthermore, a
severe shortage of mitochondrial ATP production, which presumably occurs when
the electron transfer chain is deficient, is supposedly sufficient to cause cell necrosis [7]. However, more recently, it was discovered that respiratory chain complex
II deficiency can favor cell proliferation and tumor formation rather than cell death
[8]. This has led to a renewed interest in the potential role of respiratory chain dysfunction in tumor promotion.
2 Mitochondria and Cancer: A Long History
Otto Warburg performed the seminal observation that, even under aerobic conditions, cancer cells shift their metabolism from an oxidative one (through the mitochondrial respiratory chain) to a highly glycolytic metabolism, thereby producing
elevated levels of lactate [9]. This paradoxical effect, wherein oxygen is not utilized
even though it is available, was thereafter referred to as the Warburg effect.This
phenomenon is at variance with the Pasteur effect according to which glycolysis
is naturally favored by conditions that turn off the oxygen-dependent mitochondrial respiratory chain. As a result of the shift of metabolic flux from respiration
to glycolysis, the former would become of secondary importance, and this logically suggested the idea that the fate of the mitochondria in tumor tissues would
be somewhat irrelevant to tumor development. Accordingly, up to very recently, the
word cancer was indeed largely absent from mitochondrial disease–devoted meetings and reviews. Even the reports of decreased respiratory chain activities or of
quantitative changes in mitochondrial DNA (mtDNA)-encoded mRNA in tumors
were insufficient to really draw the attention of the mitochondria-oriented scientific
community. However, several features of the mitochondrial electron transfer chain
that could play a role in triggering, or participating in, tumorigenesis deserve to be
presented.
3 The Respiratory Chain: Biochemical and Genetic Features
The respiratory chain (RC) is constituted by an association of supercomplexes
whose spatial organization favors a rapid exchange of electrons. Such a mechanism may serve to avoid—or at least to limit—direct reactions between reduced
RC carriers and molecular oxygen that can produce superoxides [10]. Two major
entities have been described: the respirasome consisting of the association of RC
complexes I/III/IV and presumably a subfraction of cytochrome c; and the ATPasome, consisting of complex V plus the adenylate and the phosphate carriers. The
The Electron Transport Chain and Carcinogenesis
21
respirasome channels the electrons from respiratory substrates to molecular oxygen to form water with a concomitant proton extrusion from the mitochondrial
matrix space to the intermembrane space (Fig. 1). This creates an electrochemical gradient across the inner membrane that can be further used by the ATPasome to generate ATP from ADP and inorganic phosphate. Additional processes
make use of the electrochemical gradient, ensuring the transport of various components through the membrane, including various cations (e.g., calcium), anions (e.g.,
organic acids, fatty acids), or proteinaceous material (e.g., mitochondrial protein
precursors) [1]. Dissipation of the electrochemical gradient can also be an aim per
se, allowing the uncoupling of electron flow from the phosphorylation process [11].
This process liberates heat and mechanically favors the oxidized state of the RC
Fig. 1 The dual genetic origin and the organization of the respiratory chain. (A) Respiratory chain
components are encoded by both the nuclear and the mitochondrial (mt) genomes. (B) A schematic
view of the organization of the respirasome and the ATPasome. CI–CV, the various complexes of
the respiratory chain; cyt c, cytochrome c; IM, inner membrane; Q: ubiquinone pools
22
J.-J. Bri´ere et al.
components, thus reducing the chance to generate superoxides. UnCoupling Proteins (UCP), a large family of proteins, mediate this process, with a high tissue
specificity. The physiologic function of a number of these UCPs is still a subject of
debate [12].
The respirasome interacts with several other dehydrogenases (e.g., complex II,
the electron transfer flavoprotein [ETF], the glycerol-3-phosphate and the dihydroorotate dehydrogenases) through the kinetically compartmentalized quinone
pool in the inner mitochondrial membrane (Fig. 1) [12]. Depending on their specific metabolic demand, tissues are variably endowed with these dehydrogenases
[13]. Redox status of the quinone pool therefore depends on the respective activities of the reducing enzymes (the various dehydrogenases) and of the oxidizing
part of the RC (i.e., the cytochromic segment). This redox status is also important
because it helps to reconstitute the reducing power of lipophilic antioxidants, such
as tocopherol (vitamin E) [14].
This complex machinery constituted of about 100 proteins is encoded by both the
mitochondrial genome (mtDNA) and the nuclear genome (Fig. 1) [15]. The maternally inherited mtDNA ranges from a few hundred (lymphocytes) to more than
100,000 (oocytes) copies per cell [16]. Each circular molecule consists of 16,569
base pairs with 37 genes encoding 13 proteins (polypeptides), 22 transfer RNA
(tRNAs), and 2 ribosomal RNAs (rRNAs) [15]. In one cell, the mtDNA copies can
be identical (a situation referred as cellular homoplasmy) or not (a situation known
as heteroplasmy). In the particular case of deleterious mutations harbored by the
mtDNA, a threshold between wild type (wt) and mutant mtDNA species has to be
reached before an RC phenotype can be observed [17]. Because of the random distribution of mitochondria during cell division, an individual can have variable levels
of heteroplasmy from tissue to tissue. The involvement of a given tissue containing mutant mtDNA will therefore depend on the actual level of heteroplasmy, the
threshold value required for RC impairment, and the susceptibility of the tissue to
the many consequences of an RC dysfunction (e.g., ATP shortage, superoxide overproduction, metabolic blockade, etc.).
The remaining components of the RC are encoded by genes present in the nuclear
genome. According to the theory of the symbiotic origin of the mitochondria, these
genes have supposedly left the mitochondria to be inserted in the nuclear chromosomes during evolution [18]. Noticeably, nonfunctional copies of genes encoding mitochondrial proteins are frequently encountered in the nuclear genome [19].
This also stands true for genes still harbored by the mtDNA, a feature that should
be carefully recorded when studying these genes [20]. Accordingly, the authenticity of the increased levels of mutant mtDNA frequently reported in tumor tissues
has been recently questioned [21]. Finally, if as mentioned earlier, the RC does
much more than to only transfer electrons and synthesize ATP, its functioning also
requires much more than the hundreds of genes that encode its constituents. Indeed
about 1500 genes on the 30,000 genes of the human genome are believed to be
necessary for the sustained building, function and maintenance of mitochondrial
functions.
The Electron Transport Chain and Carcinogenesis
23
4 Mutations in Three Nuclear Genes Encoding Respiratory
Chain Complex II Can Promote Tumor Formation
Mutations in three of the four genes encoding complex II have been shown to
cause tumor formation and cancer. Mutations of succinate dehydrogenase (SDH)
are frequently encountered in patients with an apparently sporadic form of paragangliomas (PGLs) or pheochromocytomas (PHEOs) [22]. PGLs are neuroendocrine
tumors that may secrete catecholamines [23]. They occur most frequently in the
head (glomus tympanicum and jugulare), neck (carotid body and glomus vagale),
adrenal medulla, and extra-adrenal sympathetic ganglia. PGLs are generally benign,
highly vascularized tumors occurring close to the major blood vessels and cranial
nerves [24]. About 10% of these tumors are malignant. PGLs are usually diagnosed
on the basis of the presence of a mass in the neck, pulsatile tinnitus, or hearing
loss. According to current nomenclature, the term pheochromocytomas should be
restricted to secreting PGLs evolving from the adrenal medulla, whereas a functional PGL refers to an extra-adrenal secreting PGL, which was previously known
as an ectopic PHEO. The tumors now described as carotid body PGL were formerly
known as chemodectoma due to the chemoreceptor function of the carotid body.
PGLs seem to be inherited in about 30% of cases. The hereditary form of PGL is
usually characterized by an early onset and a more severe presentation than that of
the sporadic form. These tumors often display bilateral and multiple locations and
may be recurrent or malignant. The presence of one or several secreting tumors is
not rare and contributes to the severity of inherited PGL [23]. Such PGLs constitute a genetically heterogeneous group of diseases as four different loci have been
implicated: PGL1 on 11q23, PGL2 on 11q13, PGL3 on 1q21, and PGL4 on 1p36. In
2000, linkage analysis and positional cloning allowed Baysal et al. [8] to report the
first deleterious mutations in the SDHD gene, corresponding with the PGL1 locus.
The use of a candidate gene approach has led to the identification of mutations in
SDHC (PGL3) and SDHB (PGL4) [24, 25]. Several mutations in these three genes
were subsequently reported in patients with hereditary PGL and in apparently sporadic cases [26]. These observations fueled a vigorous debate on the mechanism
linking SDH deficiency to tumor formation.
Based on previous findings on the potential consequences of a RC blockade, it
has been suggested that superoxide overproduction may be at the origin of increased
cell proliferation [27]. Accordingly, the mev1 mutant of the worm Caenorhabditis
elegans defective in the cytochrome b subunit of CII was found to have a reduced lifespan that was attributed to overproduction of superoxides [28]. In the minds of many
investigators, the superoxide overproduction provided a simple explanation linking
RC defects to neoplastic transformation. However, no hyperplasia or indications of
abnormal cell proliferation were reported in the mev1 mutant at variance with some
other C. elegans mutants for proteins that are prone to trigger tumorigenesis when
mutated in the homologous protein in mammals (e.g., cul1 mutant) [29].
Soon after the discovery that SDH mutations can result in tumor formation, it
was shown in a family harboring a SDHD mutation that these tumors were highly
24
J.-J. Bri´ere et al.
vascularized and that this corresponded with hypoxia-inducible factor (HIF) stabilization and the activation of the hypoxia pathway [30]. HIFs are composed of two
subunits: an ␣ subunit, which is regulated by oxygen, and a ␤-subunit (HIF-1␤, also
called the aryl hydrocarbon receptor nuclear translocator, or ARNT), which is constitutively and ubiquitously expressed. Three ␣-subunits have been identified to date
(HIF-1␣, HIF-2␣, and HIF-3␣), each encoded by a different gene. Under normoxic
conditions, HIF-␣ is continuously ubiquitinated and subsequently degraded by the
proteasome [31]. The process of ubiquitination is started by their recognition by the
von Hippel–Lindau (VHL) protein, which in turn requires the hydroxylation of two
proline residues on HIF-␣ [32]. The very first step of HIF-␣ degradation under normoxic conditions is thus dependent on this hydroxylation, which is catalyzed by HIF
prolyl hydroxylases (PHDs). PHDs belong to the superfamily of Fe(II)-dependent
oxygenases and require reduced iron as a cofactor, alpha-ketoglutarate (␣-KG) and
oxygen as cosubstrates, with carbon dioxide and succinate being the products of the
reaction [33]. Under hypoxic conditions, the absence of oxygen prohibits PHD activity, and HIF-␣ is thus stabilized, allowing for its nuclear translocation and the subsequent activation of its target genes. The involvement of HIFs has been observed
in numerous types of tumors, playing an active role in the progression of neoplasia
[34]. Besides oxygen, others factors, which intervene in the reaction catalyzed by
the PHD enzyme, can also affect its activity. Consistent with this hypothesis, it was
shown that elevated levels of succinate, the product of the PHDs reaction, can result
in the blockade of its activity with the consequent stabilization of the HIF-1␣ protein
(Fig. 2) [35, 36]. A significant accumulation of succinate was found in SDH-deficient
cells and tumors. Moreover, the addition of ␣-ketoglutarate, the substrate of the PHD,
Fig. 2 The interaction
between the mitochondrion
and type II hexokinase. The
channeling of mitochondrial
ATP by type II hexokinase
favors the production of
glucose 6-phosphate (G 6-P)
and ultimately glycolysis.
Ant, adenylate carrier; HKII,
type II hexokinase; I–V, the
various complexes of the
respiratory chain; Q,
ubiquinone; VDAC,
voltage-dependent anion
channel
The Electron Transport Chain and Carcinogenesis
25
was shown to abolish the nuclear translocation of HIF-1␣ in the nucleus of SDHdefective cells [35]. This provided an alternative explanation linking SDH deficiency
to tumor formation and an explanation of the restriction of the induction of tumors to
RC complex II blockade. Finally, the respiration of SDH-defective cells was shown
to be not significantly decreased indicating that the electron flow to oxygen and the
phosphorylation process are not significantly affected [37].
Additional support in favor of this latter hypothesis came from the observation
that fumarase defect can lead to leiomyomatosis and renal cell cancer (HLRCC) syndrome [38]. In this latter case, fumarate, accumulated because of fumarase inactivation, was found to act as a competitive inhibitor of the PHD, thus inducing the abnormal stabilization of HIF-1␣. Noticeably, other structurally related organic acids can
also inhibit PHD [39]. Therefore, a tricarboxylic acid cycle (TCA) cycle blockade
may result in the induction of angiogenesis and in the upregulation of glycolysis
during tumorigenesis, being at the origin of the Warburg effect. These observations
support the view that rather than a blockade of the electron flow (potentially associated with all subtypes of RC defects), the impairment of a particular segment of the
Krebs cycle can be at the origin of tumor formation [40]. It would therefore be at
least rather imprudent to invoke SDH mutations as a general proof that a RC defect
results in tumor formation.
5 Does a Defective Respiratory Chain Trigger Cell Proliferation?
To date, there is no conclusive evidence that a perturbation of the electron flow
through the RC is sufficient to increase cell proliferation and to constitute the primary cause of tumor formation. On one hand, none of the nuclear genes encoding RC components or involved in the building or maintenance of the chain have
been demonstrated to be tumor suppressors, with the exception of genes encoding RC complex II (or fumarase; see above) [41]. Noticeably, the accumulation
of ␣-ketoglutarate is a feature frequently associated with RC dysfunction, which
will not favor HIF-1␣ stabilization through PHD inhibition. On the other hand,
RC-specific poisons are not known as carcinogenic (e.g., the complex I–specific
inhibitor rotenone shown to possibly induce parkinsonism in mammals). Finally,
patients harboring deleterious mutation in genes encoding RC components are not
known to be particularly at risk for tumor formation. This is even true for mutations
affecting RC proteins such as ATPase (complex V) components that result in high
levels of superoxide production [42]. These observations lend less credibility to the
idea that an impairment of the RC function could per se be at the origin of cell
proliferation and tumor formation.
Even more puzzling is the fact that in a subset of tumor cells (liver and pancreatic
tumors), the enhanced glycolysis might well require an active production of ATP
by the RC. The hexokinase specifically expressed in most tumor cells uses both
ATP and glucose to produce ADP and glucose-6-phosphate. This tumor-specific
hexokinase is characterized by its poor affinity (high Km , about 10 mM) for glucose, similar to the hexokinase IV (glucokinase) found in normal hepatocytes [43].
26
J.-J. Bri´ere et al.
Fig. 3 Induction of the hypoxia path by succinate through HIF-1␣ stabilization. Upon succinate
dehydrogenase (complex I) blockade, succinate accumulates and is exported to the cytosol. There,
it inhibits the prolyl hydroxylase (PHD) triggering HIF-1␣ stabilization. The nuclear translocation
of this latter factor induces an increased transcription of the hypoxia pathway components. VHL,
von Hippel–Lindau; I–V, the various respiratory chain complexes
The Electron Transport Chain and Carcinogenesis
27
It possibly allows for an additional capacity for glucose uptake from plasma and
increased glucose phosphorylation by displacement of the cell glucose equilibrium.
Interestingly, whereas the isoform with a low affinity is expressed in most tumor
cells, in the tumoral liver, similarly to tumoral pancreatic cells, it is largely replaced
by a high-affinity form (hexokinase II; HKII). This latter form can readily bind
voltage-dependent anion channel (VDAC) at the outer mitochondrial membrane and
utilize the mitochondrial ATP to produce glucose-6-phosphate thus favoring an aerobic glycolysis [44] so long as the electron transfer chain is functional (Fig. 3).
Any significant impairment of RC activity would then tend to decrease the ATP
produced by the RC and oppose the upregulation of glycolysis. In addition, it has
been suggested that, upon binding to VDAC, HKII would be protected from product
inhibition because of a conformational change and in turn may reduce the chance of
VDAC to open and to facilitate release of proapoptotic factors [43, 45]. In this case,
not only would carcinogenesis not be triggered by mitochondrial dysfunction, but
also it would require the preservation of the respiratory chain function.
Finally, there is ample evidence that an impairment of RC function under most
conditions leads to cell death, tissue necrosis, and the related wide range of clinical
symptoms observed in patients harboring mutations in genes necessary for the synthesis of the electron transfer chain [46]. Potential overproduction of superoxides
has been shown to be responsible for triggering apoptotic cell death, even if low
amount of superoxides might be required for normal cell division, and can target
various cell cycle checkpoints.
It therefore appears that a defective RC is probably not per se a primary cause of
tumor formation. On the contrary, it could in some cases even constitute an obstacle
to this process!
6 Mitochondria, Superoxides, Hypoxia
The possibility nevertheless exists that low oxygenation in tumors may secondarily affect mitochondrial functioning favoring the formation of superoxides and
peroxides by the RC (Fig. 4). In principle, activated oxygen species may in turn
upregulate both oncogene growth factors and their tyrosine kinase receptors thus
driving cell transformation [47], simultaneously promoting HIF-1␣ stabilization
by inhibiting prolyl hydroxylase. In any event, superoxide production that may
promote tumorigenesis would have to be restricted to a very narrow window of
concentration, large overproduction readily favoring cell death rather than proliferation. The suggested role of superoxides in triggering tumorigenesis has long
been advocated in support for a therapeutic use of antioxidants [48]. Actually, a
wide range of antioxidant molecules is available that can potentially interfere with
cell free radicals, possibly released by mitochondria. However, contrasting results
from in vitro and in vivo (intervention trials) experiments have raised some doubt
about the ability of antioxidants to actually fight cancer by such a mechanism.
Indeed, most antioxidant molecules also act as prooxidants; this possibly accounts
28
J.-J. Bri´ere et al.
Fig. 4 Linking electron transfer chain to cell death and proliferation. (A) According to a first
hypothesis, the level of mitochondrial superoxide production largely determines the fate of the
cells. A low superoxide production favors cell division, whereas increasing this production triggers dysplastic lesion and tumor formation. An excessive production rather results in cell death.
(B) According to this second scheme, the blockade of the prolyl hydroxylase by organic acids
(succinate, fumarase) results in HIF-1␣ stabilization and the activation of the hypoxia path allowing for cell proliferation. Release of mitochondrial proapoptotic factors (AIF, cytochrome c) and
overproduction of superoxide (blockade of the ATPase; complex V) result in cell death
for their potential antitumoral activity. For instance, an antioxidant molecule like
melatonin would exercise its antiproliferative effect on the growth of rat pituitary
prolactin-secreting tumor cells in vitro by damaging mitochondria rather than by
quenching superoxides [49]. Thus even if an increased superoxide production is
observed in a subset of cancer cells, more evidence is needed to establish that,
as a general rule, these superoxides are instrumental in triggering tumorigenesis. Recent studies however have established that overexpression of thioredoxin,
manganese-dependent superoxide dismutase (MnSOD), glutathione peroxidase, or
catalase all tend to reduce cell proliferation in different animal or cell models
The Electron Transport Chain and Carcinogenesis
29
suggesting that handling of superoxides or their derivatives might be critical at some
point in a subset of cases [50, 51] (see, however, Refs. 52, 53].
7 Targeting the Respiratory Chain to Kill Tumors
Although it is not clear that a defective respiratory chain actually favors tumor formation, mitochondria might represent the Achilles’ heel of cancer cells (Fig. 5).
Indeed, mitochondria house several proapoptotic factors that are simultaneously
Fig. 5 Mitochondria and apoptosis. Release of intermembrane space components (such as AIF
or cytochrome c) and/or the loss of membrane potential (⌬⌿) under the effect of numerous stimuli trigger cell commitment to die through an apoptotic process. Loss of membrane potential and
of various matrix cofactors can result from the opening of the mitochondrial permeability transition pore (PTP). The balance between the proapoptotic and antiapoptotic members of the Bcl-2
family controls the opening of the pore. Alternatively, the members of this family can form channels that may also allow for the release of proapoptotic components present in the intermembrane
space. PTP would be a complex formed between the voltage-dependent anion channel (VDAC;
outer membrane) and the adenylate carrier (ANT; inner membrane), associated with several additional proteins. AIF, apoptosis inducing factor; Br, benzodiazepine receptor; CI–CV, the various
complexes of the respiratory chain; c, cytochrome c; IM, inner membrane; OM, outer membrane
30
J.-J. Bri´ere et al.
components of (or closely associated with) the electron transfer chain, and targeting tumors with reagents that induce release of these components has became a
fashionable idea [54]. Accordingly, cisplatin, one of the most important chemotherapeutic agents ever developed, has been shown to readily interact with mitochondria
to trigger apoptosis [55]. Resveratrol, a natural polyphenolic antioxidant, has been
reported to possess a cancer chemopreventive potential that has been attributed to
its ability to trigger mitochondrial dysfunction and cell apoptosis [56]. Actually,
an exploding list of promising therapeutic and preventative agents is being shown
(or has been suggested) to actually act through a quite similar mechanism, with or
without the involvement of superoxide production.
8 Conclusion
Mutations in both nuclear and mitochondrial genes encoding respiratory chain components have been suggested to be causative for triggering cancer. However, despite
the numerous mtDNA mutations reported in many tumors, only germ-line mutations in the complex II–encoding genes (SDHB,C,D) have been conclusively shown
as the primary cause of tumor formation. Interestingly enough, no inherited mtDNA
mutation has been observed to cause tumor to date, contrasting the high frequency
of mutant mtDNA in various type of tumors (see, however, Ref. 21). Although these
may have a role in the oncogenic process at some point, they may be a secondary
event totally irrelevant for this process. As a general rule, it seems that targeting
respiratory chain function, except for complex II activity, provides a mechanism to
inhibit rather than to promote tumor formation.
References
1. Tzagoloff A. Mitochondria. New York: Plenum Press, 1982.
2. Green DR, Kroemer G. The pathophysiology of mitochondrial cell death. Science 2004;
305(5684):626–629.
3. Krippner A, Matsuno-Yagi A, Gottlieb RA, Babior BM. Loss of function of cytochrome c in
Jurkat cells undergoing fas-mediated apoptosis. J Biol Chem 1996; 271(35):21629–21636.
4. Green DR, Reed JC. Mitochondria and apoptosis. Science 1998;281(5381):1309–1312.
5. Vahsen N, Cande C, Briere JJ, et al. AIF deficiency compromises oxidative phosphorylation.
EMBO J 2004; 23(23):4679–4689.
6. Garrido C, Kroemer G. Life s smile, death s grin: vital functions of apoptosis-executing proteins. Curr Opin Cell Biol 2004; 16(6):639–646.
7. Golstein P, Kroemer G. Cell death by necrosis: towards a molecular definition. Trends
Biochem Sci 2007; 32(1):37–43.
8. Baysal BE, Ferrell RE, Willett-Brozick JE, et al. Mutations in SDHD, a mitochondrial complex II gene, in hereditary paraganglioma. Science 2000; 287(5454):848–851.
9. Warburg O, Wind F, Negelein E. The metabolism of tumors in the body. J. General Physiol
1926; 8(6):519.
10. Schagger H, Pfeiffer K. Supercomplexes in the respiratory chains of yeast and mammalian
mitochondria. EMBO J 2000; 19(8):1777–1783.
11. Pecqueur C, Couplan E, Bouillaud F, Ricquier D. Genetic and physiological analysis of the
role of uncoupling proteins in human energy homeostasis. J Mol Med 2001; 79(1):48–56.
The Electron Transport Chain and Carcinogenesis
31
12. Gutman M. Electron flux through the mitochondrial ubiquinone. Biochim Biophys Acta 1980;
594(1):53–84.
13. Pette D, Klingenberg M, Buecher T. Comparable and specific proportions in the mitochondrial
enzyme activity pattern. Biochem Biophys Res Commun 1962; 7:425–429.
14. Geromel V, Darin N, Chretien D, et al. Coenzyme Q(10) and idebenone in the therapy
of respiratory chain diseases: rationale and comparative benefits. Mol Genet Metab 2002;
77(1–2):21–30.
15. Larsson NG, Clayton DA. Molecular genetic aspects of human mitochondrial disorders. Annu
Rev Genet 1995; 29:151–178.
16. Chen X, Prosser R, Simonetti S, Sadlock J, Jagiello G, Schon EA. Rearranged mitochondrial
genomes are present in human oocytes. Am J Hum Genet 1995; 57(2):239–247.
17. Bourgeron T, Chretien D, Rotig A, Munnich A, Rustin P. Fate and expression of the deleted
mitochondrial DNA differ between human heteroplasmic skin fibroblast and Epstein-Barr
virus-transformed lymphocyte cultures. J Biol Chem 1993; 268(26):19369–19376.
18. Rustin P, Jacobs HT, Dietrich A, Lightowlers RN, Tarassov I, Corral-Debrinski M. Targeting allotopic material to the mitochondrial compartment: new tools for better understanding
mitochondrial physiology and prospect for therapy. Med Sci (Paris) 2007; 23(5):519–525.
19. Ricchetti M, Tekaia F, Dujon B. Continued colonization of the human genome by mitochondrial DNA. PLoS Biol 2004; 2(9):E273.
20. Parfait B, Rustin P, Munnich A, Rotig A. Co-amplification of nuclear pseudogenes and assessment of heteroplasmy of mitochondrial DNA mutations. Biochem Biophys Res Commun
1998; 247(1):57–59.
21. Salas A, Yao YG, Macaulay V, Vega A, Carracedo A, Bandelt HJ. A critical reassessment of
the role of mitochondria in tumorigenesis. PLoS Med 2005; 2(11):e296.
22. Briere JJ, Favier J, Gimenez-Roqueplo AP, Rustin P. Tricarboxylic acid cycle dysfunction as
a cause of human diseases and tumour formation. Am J Physiol Cell Physiol 2006; 291(6):
1114–1120.
23. Favier J, Briere JJ, Strompf L, et al. Hereditary paraganglioma/pheochromocytoma and inherited succinate dehydrogenase deficiency. Horm Res 2005; 63(4):171–179.
24. Astuti D, Latif F, Dallol A, et al. Gene mutations in the succinate dehydrogenase subunit
SDHB cause susceptibility to familial pheochromocytoma and to familial paraganglioma. Am
J Hum Genet 2001; 69(1):49–54.
25. Niemann S, Muller U. Mutations in SDHC cause autosomal dominant paraganglioma, type 3.
Nat Genet 2000; 26(3):268–270.
26. Nakamura E, Kaelin WG Jr. Recent insights into the molecular pathogenesis of pheochromocytoma and paraganglioma. Endocr Pathol 2006; 17(2):97–106.
27. Rustin P. Mitochondria, from cell death to proliferation. Nat Genet 2002; 30(4):352–353.
28. Senoo-Matsuda N, Yasuda K, Tsuda M, et al. A defect in the cytochrome b large subunit in
complex II causes both superoxide anion overproduction and abnormal energy metabolism in
Caenorhabditis elegans. J Biol Chem 2001; 276(45):41553–41558.
29. Piva R, Liu J, Chiarle R, Podda A, Pagano M, Inghirami G. In vivo interference with Skp1
function leads to genetic instability and neoplastic transformation. Mol Cell Biol 2002;
22(23):8375–8387.
30. Gimenez-Roqueplo AP, Favier J, Rustin P, et al. The R22X mutation of the SDHD gene in
hereditary paraganglioma abolishes the enzymatic activity of complex II in the mitochondrial
respiratory chain and activates the hypoxia pathway. Am J Hum Genet 2001; 69(6):1186–1197.
31. Hickey MM, Simon MC. Regulation of angiogenesis by hypoxia and hypoxia-inducible factors. Curr Top Dev Biol 2006; 76:217–257.
32. Kaelin WG. Proline hydroxylation and gene expression. Annu Rev Biochem 2005;74:
115–128.
33. Lee JW, Bae SH, Jeong JW, Kim SH, Kim KW. Hypoxia-inducible factor (HIF-1)alpha: its
protein stability and biological functions. Exp Mol Med 2004; 36(1):1–12.
34. Gordan JD, Simon MC. Hypoxia-inducible factors: central regulators of the tumor phenotype.
Curr Opin Genet Dev 2007; 17(1):71–77.
32
J.-J. Bri´ere et al.
35. Briere JJ, Favier J, Benit P, et al. Mitochondrial succinate is instrumental for HIF1alpha
nuclear translocation in SDHA-mutant fibroblasts under normoxic conditions. Hum Mol
Genet 2005; 14(21):3263–3269.
36. Selak MA, Armour SM, MacKenzie ED, et al. Succinate links TCA cycle dysfunction to
oncogenesis by inhibiting HIF-alpha prolyl hydroxylase. Cancer Cell 2005; 7(1):77–85.
37. Bourgeron T, Rustin P, Chretien D, et al. Mutation of a nuclear succinate dehydrogenase gene
results in mitochondrial respiratory chain deficiency. Nat Genet 1995; 11(2):144–1449.
38. Tomlinson IP, Alam NA, Rowan AJ, et al. Germline mutations in FH predispose to dominantly
inherited uterine fibroids, skin leiomyomata and papillary renal cell cancer. Nat Genet 2002;
30(4):406–410.
39. MacKenzie ED, Selak MA, Tennant DA, et al. Cell-permeating alpha-ketoglutarate derivatives alleviate pseudohypoxia in succinate dehydrogenase-deficient cells. Mol Cell Biol 2007;
27(9):3282–3289.
40. Briere JJ, Favier J, El Ghouzzi V, et al. Succinate dehydrogenase deficiency in human. Cell
Mol Life Sci 2005; 62(19–20):2317–2324.
41. Kroemer G. Mitochondria in cancer. Oncogene 2006; 25(34):4630–4632.
42. Geromel V, Kadhom N, Cebalos-Picot I, et al. Superoxide-induced massive apoptosis in cultured skin fibroblasts harboring the neurogenic ataxia retinitis pigmentosa (NARP) mutation
in the ATPase-6 gene of the mitochondrial DNA. Hum Mol Genet 2001; 10(11):1221–1228.
43. Brandon M, Baldi P, Wallace DC. Mitochondrial mutations in cancer. Oncogene 2006;
25(34):4647–4662.
44. Bustamante E, Pedersen PL. High aerobic glycolysis of rat hepatoma cells in culture: role of
mitochondrial hexokinase. Proc Natl Acad Sci USA 1977; 74(9):3735–3739.
45. Mathupala SP, Ko YH, Pedersen PL. Hexokinase II: cancer s double-edged sword acting as
both facilitator and gatekeeper of malignancy when bound to mitochondria. Oncogene 2006;
25(34):4777–4786.
46. Munnich A, Rustin P. Clinical spectrum and diagnosis of mitochondrial disorders. Am J Med
Genet 2001; 106(1):4–17.
47. Aslan M, Ozben T. Oxidants in receptor tyrosine kinase signal transduction pathways.
Antioxid Redox Signal 2003; 5(6):781–788.
48. Nishikawa M, Hashida M. Inhibition of tumour metastasis by targeted delivery of antioxidant
enzymes. Expert Opin Drug Deliv 2006; 3(3):355–369.
49. Yang QH, Xu JN, Xu RK, Pang SF. Antiproliferative effects of melatonin on the growth of rat
pituitary prolactin-secreting tumor cells in vitro. J Pineal Res 2007; 42(2):172–179.
50. Weydert CJ, Waugh TA, Ritchie JM, et al. Overexpression of manganese or copper-zinc superoxide dismutase inhibits breast cancer growth. Free Radic Biol Med 2006; 41(2):226–237.
51. Li GX, Hirabayashi Y, Yoon BI, et al. Thioredoxin overexpression in mice, model of attenuation of oxidative stress, prevents benzene-induced hemato-lymphoid toxicity and thymic
lymphoma. Exp Hematol 2006; 34(12):1687–1697.
52. Lu YP, Lou YR, Yen P, et al. Enhanced skin carcinogenesis in transgenic mice with high
expression of glutathione peroxidase or both glutathione peroxidase and superoxide dismutase. Cancer Res 1997; 57(8):1468–1474.
53. Welsh SJ, Bellamy WT, Briehl MM, Powis G. The redox protein thioredoxin-1 (Trx-1)
increases hypoxia-inducible factor 1alpha protein expression: Trx-1 overexpression results
in increased vascular endothelial growth factor production and enhanced tumor angiogenesis.
Cancer Res 2002; 62(17):5089–5095.
54. Galluzzi L, Larochette N, Zamzami N, Kroemer G. Mitochondria as therapeutic targets for
cancer chemotherapy. Oncogene 2006; 25(34):4812–4830.
55. Cepeda V, Fuertes MA, Castilla J, Alonso C, Quevedo C, Perez JM. Biochemical mechanisms
of cisplatin cytotoxicity. Anticancer Agents Med Chem 2007; 7(1):3–18.
56. Fulda S, Debatin KM. Resveratrol modulation of signal transduction in apoptosis and cell
survival: a mini-review. Cancer Detect Prev 2006; 30(3):217–223.
Respiratory Control of Redox Signaling
and Cancer
Pauline M. Carrico, Nadine Hempel and J. Andr´es Melendez
Abstract In the past several decades, elegant work has been performed in a variety
of organisms suggesting that decreases in oxidant burden can prolong life span and
decrease the severity of age-associated degenerative disorders (Finkel T, Holbrook
NJ. Nature 2000; 408:239–247). Cancer is primarily an age-related disease and
is associated with the altered production of oxidants (Cerutti PA. Science 1985;
227:375–381). Compounds that generate reactive oxygen species (ROS) promote
skin tumors in mice. Treatment with antioxidants that terminate the chain reactions
initiated by ROS antagonizes this process. Many tumor cell types have also been
shown to have an increased oxidant radical load relative to normal tissue. Thus,
tumor cells have adapted to cope with an increased oxidant burden compared with
that of normal nonmalignant tissue and may even harness the increased oxidant
load to modulate signaling pathways that are critical for their survival. With the
tight connection between oxidant production and tumor promotion and growth, it
is quite surprising that antioxidant-based therapeutics have not become pervasive in
treatments for many cancers (for a concise review on this topic, see Seifried HE,
et al. Cancer Research 2003; 63:4295–4298).
Hydrogen peroxide (H2 O2 ) is the electron-neutral, 2e- reduction product of oxygen that is generated both chemically and enzymatically and can modulate the
activity of a variety of signaling molecules. This chapter will summarize how the
metabolic production of H2 O2 can impact distinct signaling pathways that are essential for successful tumorigenesis and metastatic progression and will identify redoxsensitive pathways that may be targets for antioxidant-based cancer therapy.
Keywords Reactive oxygen species · Cancer · Antioxidants · Redox signaling ·
Hydrogen peroxide
J.A. Melendez (B)
Center for Immunology and Microbial Disease MC151, Albany Medical College, Albany, NY
12208
e-mail: melenda@mail.amc.edu
S.P. Apte, R. Sarangarajan (eds.), Cellular Respiration and Carcinogenesis,
DOI 10.1007/978-1-59745-435-3 3,
C Humana Press, a part of Springer Science+Business Media, LLC 2009
33
34
P.M. Carrico et al.
1 ROS and Redox Signaling
All aerobic organisms generate ROS as by-products of normal cellular metabolism.
Although the concentrations of ROS are normally low but measurable, accumulation
of these species can be toxic to the cells. To protect themselves from the deleterious
effects of ROS, cells have evolved an impressive repertoire of enzymatic and nonenzymatic antioxidant defenses. In normal cells, a balance between the rates of ROS
production and the scavenging capacity by antioxidants is required to maintain the
cellular redox homeostasis. An imbalance resulting from either an increase in ROS
concentration or a decrease in the antioxidant capacity leads to the modulation of
redox-dependent signaling events.
ROS production can arise from a variety of sources in nonphagocytic cells. Most
ROS are produced as by-products of the electron transport chain of the mitochondria at the level of complex I (NADH/ubiquinone oxidoreductase) and complex III
(ubiquinol/cytochrome c oxidoreductase). Outside the mitochondria, ROS can originate from enzymes such as NADPH oxidase, dual oxidases, and lipoxygenases.
At the cell surface, ROS production can be initiated after a variety of stimuli, such
as integrin signaling, growth factors, and cytokines, such as TNF-␣, which activate NADPH oxidase. ROS can also arise from arachidonic acid metabolism via
5-lipoxygenase. Both cytoplasmic and mitochondrial pathways create the highly
unstable superoxide anion (O2 -. ), which is either chemically or enzymatically converted to H2 O2 . The enzyme responsible for this dismutation is superoxide dismutase, isoforms of which can be found in the cytoplasm (copper zinc/sod1) or the
mitochondria (manganese /sod2).
Conditions where mitochondrial-derived ROS are produced at a rate that exceeds
their scavenging capacity are commonly associated with metabolic perturbations
that impact signaling. Recent evidence also suggests that ROS produced as a result
of normal metabolic flux affect signaling. For example, Nemoto et al. [1] demonstrated that the addition of extracellular pyruvate to HeLa cells resulted in a dramatic
increase in the phosphorylation/activation of JNK1, a modest increase in ERK activity, and no effect on JNK2 or p38. This activation was accompanied by an increase
in mitochondrial oxidants and was reversible by the addition of N-acetylcysteine
(NAC), overexpression of glutathione S-transferase, or extracellular catalase. These
data suggest that the addition of pyruvate results in an increase in H2 O2 levels and
the subsequent phosphorylation of JNK1. Furthermore, the activation of JNK1 coincided with the inhibition of glycogen synthase kinase 3-␤ and glycogen synthase.
This work provided convincing evidence that ROS produced as a result of normal
cellular metabolism could act as a regulator of signal transduction.
The principle mediator of ROS-dependent signaling is H2 O2 . H2 O2 generated in
response to receptor stimulation has been shown to be an efficient signal transducing
molecule by its ability to reversibly oxidize active site cysteines [2]. Many protein
tyrosine phosphatases (PTPs) are particularly susceptible to H2 O2 -dependent inactivation because of the lowered pKa of the active site cysteine. The essential cysteine residue in the PTP signature active-site motif Cys-(X)5 -Arg exists as a thiolate
anion (Cys-S- ), which, at neutral pH, is susceptible to nucleophilic attack by H2 O2 .
Respiratory Control of Redox Signaling and Cancer
35
Oxidation of the active-site cysteine generates a sulfenic derivative (Cys-SOH)
leading to enzyme inactivation that can be reversed by cellular thiols. This mode of
inactivation is similar to that described for both bacterial and eukaryotic peroxiredoxin family members that are responsible for detoxification of organic hydroperoxides, H2 O2 and ONOO- [3]. A unique feature of the eukaryotic peroxiredoxins
(PRXs) is their sensitivity to inactivation by H2 O2 levels that exceed steady state.
This sensitivity to inactivation has led to the “floodgate” hypothesis, whereby PRX
inactivation can lead to a rapid rising burst of signaling H2 O2 , which may lead
to PTP inactivation and subsequent enhancement of kinase signaling [4]. Several
PTP family members have been shown to be oxidatively inactivated by physiologic stimuli in vivo including the dual lipid protein phosphatase and tensin homolog
deleted on chromosome 10 (PTEN), the low-molecular-weight PTPs (LMW-PTPs),
and the mitogen activated protein (MAP) kinase phosphatase family members [5].
Such redox-based signaling, arising from a chronic shift in ROS homeostasis that
results from metabolic changes or altered antioxidant scavenging, can impact critical
components of the malignant phenotype. Tumor cell proliferation, survival, vascular recruitment, and migration are components that drive the malignant phenotype.
Below we discuss how alterations in metabolic ROS production can regulate these
mechanisms.
2 Mitochondrial ROS and Cell Cycle
ROS have been shown to be potent mitogens. Effective detoxification of H2 O2 by
catalase inhibits proliferation in Her-2/neu–transformed Rat-1 fibroblasts. Similarly,
addition of exogenous H2 O2 at low concentrations and its intracellular production
by the noninflammatory NADPH oxidase family members (Nox1-5) is also mitogenic [6].
H2 O2 may directly induce cell proliferation, however, it also acts as a downstream effector of platelet-derived growth factor (PDGF)-induced mitogenic signaling. In Rat-1 cells treated with PDGF, the SH2 domain–containing PTP, SHP-2,
can be oxidatively inactivated, resulting in the phosphorylation and activation of
MAPK. Both the inactivation of SHP-2 and resulting activation of MAPK is abrogated by treatment of cells with the H2 O2 scavenger NAC, demonstrating a regulatory role of H2 O2 in cell proliferation through the reversible oxidation of PTPs
[7]. In this context, antioxidant strategies can attenuate the proliferative capacity of
numerous tumor cell lines [8]. H2 O2 detoxification by the use of adenoviral catalase can effectively inhibit DNA synthesis and decrease the proliferation of human
aortic endothelial cells. Goswami and co-workers have demonstrated that intracellular H2 O2 is also required for G1-S transit in mouse embryonic fibroblasts. Thus,
in contrast with the growth inhibition commonly associated with the bolus addition
of H2 O2 , the intracellular production of submicromolar concentrations of H2 O2 is
mitogenic and stimulates proliferation, and its effective removal can prevent cell
cycle progression. It is likely that the intracellular redox state fluctuates as a result
of alterations in production or removal of mitochondrial-derived ROS. Menon and
36
P.M. Carrico et al.
Goswami have proposed that periodic fluctuations in metabolic ROS production
likely control cell proliferation via redox-sensitive cell cycle regulatory proteins [9].
The rapid turnover of antioxidants may give rise to acute shifts in cellular oxidant
production that can control normal cell cycle progression. The elucidation of oxidant scavenging or oxidant generating enzymes whose levels fluctuate with respect
to cell cycle would give credence to this hypothesis. During the process of carcinogenesis, a similar response may occur as a result of the loss of tumor suppressor
function or oncogene activation. Thus, the enhanced proliferative capacity of tumor
cells may be attributed to chronic shifts in respiratory ROS production due to perturbations in metabolic or antioxidant function.
3 Mitochondrial ROS and Cell Survival
H2 O2 has been implicated as a strong inducer of apoptosis in various cell types.
Apoptosis or programmed cell death refers to a series of tightly coordinated events
designed to eliminate potentially dangerous cells and cells that have reached the
end of their life cycle. Apoptosis allows an organism to tightly control cell number
and tissue size, as well as to protect itself from cells that interfere with homeostasis. Apoptosis is a tightly regulated process whereby a set of cysteine proteases
(caspases) are made active through a complex signaling cascade resulting in degradation of cellular nuclear DNA [10]. In contrast with necrosis, apoptosis destroys
only the cell in question with no effect on surrounding cells. Apoptosis can be triggered by both extracellular stimuli as well as intracellular events that mark cellular
dysfunction. The hallmarks of apoptosis include membrane blebbing, cytoskeletal
disorganization, and loss of mitochondrial integrity and DNA fragmentation. The
ultimate step in apoptosis is the activation of caspases, which act on a number of
downstream cellular proteins inactivating them by catalytic cleavage at an Asp-Xxx
sequence. The pro/inactive forms of these enzymes are activated via cleavage by
other upstream caspases or via autoactivation. Caspases are involved in the degradation of important cellular proteins as well as activation of caspase-activated DNAses
that cleave mitochondrial DNA, ultimately leading to cell death. Neoplastic cells are
able to evade apoptosis and circumvent cell death, an essential survival step during
tumorigenesis.
TNF-␣ is produced by cells under stress and induces signaling via binding to its
cell surface receptor. TNF-␣ along with cycloheximide (CHX) has been established
as a potent inducer of apoptosis in various tumor cell lines. One of the primary
mechanisms of TNF-␣/CHX–induced apoptosis involves activation of the receptorassociated death domains that recruit and activate caspase-8, leading to activation of
downstream caspases and finally of the executioner caspase, caspase-3. In addition,
CHX enhances the induction of apoptosis via TNF-␣ by inhibition of new protein
synthesis. One of the proteins that was first discovered to be essential for protection from TNF-mediated cell death is the mitochondrial manganese containing
superoxide dismutase (Sod2) [11]. Since this initial observation by Wong and
Respiratory Control of Redox Signaling and Cancer
37
Goeddel, numerous reports have demonstrated a role for Sod2 in protection from
many apoptotic stimuli. The antiapoptotic effects of Sod2 suggest that either the
removal of O2 -. or the generation of H2 O2 via Sod2 may have an inhibitory effect
on programmed cell death.
The mitochondrial generation of H2 O2 has been shown to confer resistance
to apoptotic stimuli. Procaspases contain an active site cysteine that is susceptible to oxidation thereby preventing their activation via proteolytic cleavage.
This may explain the ability of oxidants in some studies to deter apoptotic cell
death. Cederbaum and Bai have demonstrated that catalase overexpression in
HepG2 hepatocarcinoma cells increases their susceptibility to TNF-␣–induced
apoptosis. Fas-induced apoptotic cell death can be enhanced by overexpression
of mitochondrial-targeted catalase. The antioxidant flavonoid and terpenoid compounds derived from the Ginkgo biloba plant have been shown to enhance the
expression of proapoptotic molecules as a result of its ability to scavenge H2 O2
in distinct cancer cell types. Recent studies from this lab indicate that efficient and
mitochondrial-targeted H2 O2 detoxification can reverse the antiapoptotic properties
of Sod2 overexpression [12].
Another mechanism whereby H2 O2 may augment cell survival is via oxidative
inactivation of the LMW-PTPs. LMW-PTP activity is dependent upon the tyrosine
phosphorylation on Tyr131 and Tyr132; phosphorylation on Tyr131 results in a 25fold increase in enzymatic activity, and phosphorylation of Tyr132 serves as a scaffold for the SH2 domain of the Grb2 adaptor protein. In PDGF-treated NIH-3T3
or human prostatic carcinoma PC3 cells, inactivation of LMW-PTP by glucose oxidase (G/O)-induced oxidative stress inhibits LMW-PTP auto-dephosphorylation on
Tyr131 and Tyr132, thus enhancing the recruitment of Grb2 to LMW-PTP and, subsequently, ERK activation and cell survival [13].
Thus, even though it has been observed that bolus additions of H2 O2 induce
apoptosis, these studies have established that submicromolar concentrations of
mitochondrial-derived H2 O2 may provide a selective inhibition of programmed cell
death and that effective H2 O2 detoxification may also sensitize cancer cells to apoptotic stimuli.
4 Mitochondrial ROS and Angiogenesis
H2 O2 has been shown to regulate proangiogenic responses in a variety of systems
including human retinal pigment epithelial cells, cultured keratinocytes, and bovine
pulmonary artery endothelial cells. Monte et al. have demonstrated that H2 O2 contributes to promoting the angiogenic activity of tumor-bearing lymphocytes and that
this activity can be inhibited by coadministration of H2 O2 -detoxifying enzyme catalase but not superoxide dismutase. In vitro angiogenesis of bovine thoracic aorta can
be induced with relatively low concentrations (1 ␮M) of H2 O2 and blocked by catalase. The discovery of a noninflammatory NADPH oxidase (Nox1-5) has provided
an additional source for the production of O2 -. and H2 O2 in a variety of tumor cell
38
P.M. Carrico et al.
types [14]. The Nox1-dependent generation of H2 O2 is a potent trigger of angiogenesis, increasing the vascularity of tumors and inducing molecular markers of angiogenesis including vascular endothelial growth factor (VEGF), VEGF receptors, and
matrix metalloproteinase (MMP) activity in cultured cells and in tumors. Coexpression of catalase blocks the increased activity of the angiogenic markers, indicating that H2 O2 signals part of the switch to the angiogenic phenotype. Furthermore,
MMPs are important contributors to the angiogenic switch, and their expression is
redox-responsive [15].
The transcription factor Ets-1 is also an important regulator of angiogenesis and
is sensitive to H2 O2 production [15]. The role of Ets-1 family members in tumorigenesis is largely due to their ability to activate the transcription of a number of
genes involved in matrix degradation, such as MMP-1, 2, 3, and 9 as well as urokinase plasminogen activator. These proteases contribute to tumor invasion and the
early phases of blood vessel formation [16]. In cisplatin-resistant ovarian carcinoma
variants, elevations in H2 O2 are attributable to alterations of mitochondrial function
relative to parental cell lines leading to an increase in transcription of Ets-1. The
increase in transcription enhances binding of Nrf2, a key transcription factor in the
cellular response to oxidative stress, to an antioxidant response element (ARE) in
the promoter of Ets-1. Ets-1 expression has been correlated with poor prognosis in
breast and ovarian cancer. Furthermore, the development of cisplatin resistance is a
major limitation of this chemotherapy regime, which is widely used in the treatment
of testicular and ovarian cancer. These observations suggest that shifts in the steadystate production of H2 O2 in response to receptor engagement or altered mitochondrial function resulting from the development of cisplatin resistance can enhance
the expression of the angiogenic transcription factor Ets-1.
VEGF is a key mediator in regulating the vascular recruitment of tumor endothelial cells. The transcriptional activity of VEGF is directly responsive to H2 O2 and
involves a GC-rich region of the promoter between −95 and −51. Detailed characterization of the promoter has identified two SP1 and SP3 sites at −73/−66 and
−58/−52 that represent the core mechanism of oxidative stress–triggered VEGF
transactivation. Connor et al. have also established that mitochondrial-derived H2 O2
resulting from the dismuting activity of Sod2 was responsible for enhanced VEGFdependent angiogenic activity. The mechanism involved the H2 O2 -dependent oxidative inactivation of the dual lipid-protein phosphatase PTEN, leading to increase in
activity of the serine/threonine-specific protein kinase AKT and subsequent VEGF
expression [17].
Hypoxia inducible factor 1␣ (HIF-1␣), the master regulator of cellular responses
to hypoxia, responds to shifts in the metabolic production of ROS [18]. HIF-1␣
regulates the transcriptional activity of genes that participate in angiogenesis, iron
metabolism, glucose metabolism, and cell proliferation/survival. The regulation of
HIF-1␣ in response to hypoxia is controlled by its hydroxylation state. Hydroxylation is controlled by a family of prolyl hydroxylases (PHDs) that require molecular
oxygen for full activity. In the presence of oxygen, PHD activity is maximal, leading
to HIF-1␣ hydroxylation. Hydroxylated HIF-1␣ is then targeted for polyubiquitination leading to its degradation by the cellular proteasome complex. In the absence of
Respiratory Control of Redox Signaling and Cancer
39
O2 , hydroxylation is inhibited, leading to HIF-1␣ stabilization and activation of its
target genes. Studies have determined that efficient mitochondrial ROS scavenging
can prevent the hypoxic induction of HIF-1␣. This ROS-dependent stabilization of
HIF-1␣ involves mitochondrial complex III–dependent production of superoxide
and subsequent inhibition of PHD activity. The mechanisms for how mitochondrial
ROS regulate HIF-1␣ are not clear but may involve a shift in the oxidation state
of PHD-bound iron. It also possible that the hypoxic production of ROS may lead
to the oxidative inactivation of phosphatases, such as PTEN, that contribute to the
maximal activation of VEGF.
Thus, it appears that angiogenic signaling is particularly sensitive to alterations
in the metabolic production of oxidants and that mitochondrial free radicals may
play a role in preserving levels of their initial substrate, oxygen, via HIF-1␣.
5 Mitochondrial ROS and Invasion/Migration
A hallmark of malignancy and a primary cause of morbidity and mortality in cancer
patients is metastasis. Tumor metastasis is a multistep process initiated by migration
and invasion of cells into the surrounding vasculature. This is followed by extravasation from the circulation and proliferation at a secondary site where the formation
of a metastatic lesion occurs. A key process in the initial movement and subsequent
invasion of neoplastic cells into the circulatory system is the remodeling of the extracellular matrix (ECM). Emerging evidence has implicated ROS and the activation of
redox-sensitive signaling pathways in invasion and migration. Intrinsic antioxidant
enzymes are vital to the regulation of oxidative stress within cells. In vitro studies have shown that a number of cancer cell lines contain elevated levels of Sod2
and decreased levels of catalase and that this change in steady-state levels of H2 O2
correlates with increased metastasis, proliferation, and resistance to apoptosis.
An integral part of migration and adhesion is the dynamic rearrangement of
focal adhesion complexes in the cell. The role of Rho GTP-binding proteins (Rho,
Rac, and Cdc42) in regulating cytoskeletal dynamics by activation of Src and focal
adhesion kinase (FAK) and initiating remodeling of integrins has been firmly established. In adherent cells, mature focal adhesions have been shown to be regulated
by RhoA, whose activation results in actin stress fiber formation. Rac and Cdc42
are considered to be promigratory resulting in the formation of focal complexes of
lamellipodia and filopodia, respectively. It has been shown that ROS modulate Rac
and Rho signaling. Rac can specifically downregulate RhoA in a ROS-dependent
manner through oxidative inactivation of LMW-PTP, which normally dephosphorylates and inactivates the Rho GTPase-activating protein, p190Rho-GAP. Activation
of p190Rho-GAP results in the conversion of the GTP-bound to the GDP-bound
form of Rho, and its inactivation contributes to a loss of adhesion and a transition
to a migratory phenotype. While ROS appear to disrupt the formation of mature
focal adhesions of adherent cells, there is evidence that suggests a role of ROS in
the turnover of leading edge focal complexes observed in lamellipodia of migrating
cells. Rac has been shown to activate NADPH oxidase, which is spatially localized
40
P.M. Carrico et al.
to membrane ruffles of the leading edge of migrating cells, and was shown to be
associated with cell migration. In this context, oxidants can reversibly inactivate the
protein tyrosine phosphatase PTP-PEST that normally dephosphorylates and inactivates FAK, leading to a more rapid turnover of active FAK at focal complexes. It
is also possible that mitochondrial-derived ROS are involved in disrupting mature
focal adhesions yet contribute to active remodeling of focal complexes of the lamellipodia required for migration.
Lipid signaling molecules also play an important role in regulating the migratory phenotype. Studies in phagocytic cells and the amoeba Dictyostelium discoideum demonstrate that phosphoinositide (3,4,5) 3-phosphate (PtdIns[3,4,5]P3 )
accumulates at the front of chemotaxing cells. Concomitant with the distribution
of PtdIns(3,4,5)P3 to the leading edge is a redistribution of PTEN to the remaining
outside perimeter of the cell where it degrades PtdIns(3,4,5)P3 by removing the 3phosphate. Alterations in the steady-state levels of H2 O2 can inactivate the tumor
suppressor PTEN leading to an augmentation of AKT signaling and redistribution
of phosphoinositides at the plasma-lamellar surface [17]. It is intriguing to hypothesize that receptor engagement in response to chemotactic triggers may lead to the
local oxidant-dependent inactivation of PTEN and rapid reproportioning of lipid
signaling molecules that promote the migratory phenotype.
An essential and rate-limiting step in metastasis is the remodeling and degradation of the ECM and basement membrane by proteolytic enzymes. In this model,
cells must interact with surrounding stromal cells, leading to the loss of matrix
function and resulting in a compromised matrix boundary. MMPs are major contributors of stromal degradation and are vital to the process of cellular invasion.
MMP-1/interstitial collagenase has a specificity for type I collagen, the primary collagen of the interstitial matrix; though it has been shown to degrade collagen types
II, III, VII, and X. Studies using antisense mRNA against MMP-1 in melanoma
cells have shown a significant attenuation in the capacity of these cells to invade
type I collagen and Matrigel in vitro. Moreover, the expression of MMP-1 has been
shown to be a definitive prognostic marker for breast lesions that will develop into
cancer suggesting that the expression and regulation of MMP-1 may be important
in malignancy.
Studies from this and other laboratories have demonstrated that the Sod2dependent production of H2 O2 leads to increased expression of MMP family members and that there is a strong correlation between this increase in MMP levels and
enhanced metastasis. Interestingly, the Sod2-dependent increases in MMP expression can be reversed by the H2 O2 detoxifying enzyme, catalase, or glutathione peroxidase [15]. Thus, H2 O2 -dependent upregulation of MMPs may, in part, contribute
to increased invasion and metastatic capacity of tumors displaying elevated Sod2
levels.
Maximal mitochondrial H2 O2 -dependent transcription of MMP-1 transcription
is achieved by a proximal activator protein 1 (AP-1) site at –1602 and a single
nucleotide polymorphism (SNP) guanine insertion at –1607, which creates an Ets-1
binding site [15]. This redox-sensitive MMP-1 protein expression requires activation of both ERK1/2 and JNK pathways. JNK signaling is largely responsible for
Respiratory Control of Redox Signaling and Cancer
41
the H2 O2 sensitivity of the MMP-1 promoter, whereas ERK1/2 contributes to both
its basal and H2 O2 dependence [19].
The ability of a tumor cell to migrate and invade through the extracellular matrix
is attributed to local shifts in signaling that regulate the distribution of phospholipid signaling molecules, rearrangement of focal adhesion complexes, and matrix
degradation. All of these processes are redox-sensitive and are likely responsive
to metabolic shifts in the steady-state production of H2 O2 . As many of the processes are highly compartmentalized, it is possible that vicinal mitochondria may
participate in these migratory signaling events by augmenting their local production
of H2 O2 . With the development of fluorescent reporter molecules that detect focal
shifts in the production of many of the migratory signaling intermediates and new
redox-sensing mitochondrial targeted GFP constructs, we will be able to determine
if these events are juxtaposed.
6 Mitochondrial ROS and Cancer Stem Cells
Recently, the hypothesis has emerged that conventional therapies may fail to completely eradicate cancer because cancer stem cells, which are thought to act as
cancer-initiating cells, are insensitive to these treatments. Work in leukemia, breast,
brain, and lung cancers have provided further understanding of the role cancer stem
cells play in malignancies. Although cancer stem cells appear to have some of the
phenotypic and functional properties of embryonic stem cells, such as the potential
to give rise to a larger population of cells, survival of these two different cell types
requires different pathways that are frequently at odds with one another. Recent
research indicates that signaling by ROS may regulate growth and differentiation in
both hematopoietic stem cells and cancer stem cells.
PTEN, like p53 is one of the most common targets for mutation in sporadic
human cancers. The oxidative inactivation of PTEN and subsequent activation of the
AKT signaling pathway may lead to the transformation of adult hematopoietic stem
cells (HSCs) into leukemias. Transgenic mice with conditional deletions of PTEN in
adult HSCs develop leukemias within 4 to 6 weeks along with a depletion of normal
HSCs. These defects are dependent upon the downstream effector mammalian target
of rapamycin (mTOR), a serine/threonine kinase that regulates metabolism and that
has also been shown to generate ROS [20].
The oxidative inactivation of PTEN and subsequent activation of the AKT signaling pathway leads to the phosphorylation of the class O of forkhead transcription factors (FoxO), resulting in the retention of FoxO in the cytoplasm through
interactions with 14-3-3 proteins [21]. FoxOs, which have also been postulated to
act as tumor suppressors, regulate the transcription of genes involved in cell cycle
arrest, apoptosis, glucose metabolism, and stress response. In C. elegans, the FoxO
ortholog DAF-16 has been shown to affect life span through resistance to oxidative stress and adaptation to food shortage. DAF-16 promotes resistance to oxidative stress by activating transcription of Sod2 and catalase. In mammals, the FoxO
42
P.M. Carrico et al.
subgroup consists of four members (FoxO1, FoxO3, FoxO4, and FoxO6), of which
three (FoxO1, FoxO3, FoxO4) overlap in function and expression patterns. Individual deletion of FoxO proteins in mice does not result in a tumor-prone phenotype, however, mice containing conditional deletions of FoxO1, FoxO3, and FoxO4
display a cancer-prone condition characterized by thymic lymphomas and hemangiomas [22]. In addition, mice with conditional deletions of FoxO1, FoxO3, and
FoxO4 exhibited premature aging of HSCs, which was shown to be at least partly
due to an increase in ROS levels in the HSC compartment and attributable to a
decrease in expression of antioxidant genes [23]. Treatment of these mice with NAC
restored the HSC compartment size, reestablished stem cell cycling, and normalized
HSC apoptosis.
FOXO3a is also regulated via p66Shc, which belongs to a family of adaptors that
function in signal transduction of mitogenic and apoptotic responses, and whose
absence has been shown to prolong life span both in cell culture and mouse models. Although p66Shc is activated through tyrosine phosphorylation in response to
extracellular signals such as EGF and insulin, phosphorylation of p66Shc at serine 36 in response to oxidative stress leads to the subsequent phosphorylation and
sequestration of FOXO3a in the cytosol [24]. This may further increase levels of
ROS as FOXO3a has been shown to activate a reporter under the control of the
human catalase promoter [24].
In a hypoxic environment, p66shc functions to promote survival of stem cells
and mammary cancer stem cells through the upregulation of Notch-3 and carbonic
anhydrase IX (CA-IX) [25]. In contrast, as discussed above, inhibition of FOXO
proteins in response to H2 O2 augments apoptosis of stem cells.
Recently, p66shc was also shown to regulate mitochondrial metabolism [26]. In
immortalized mouse embryonic fibroblasts (MEFs) lacking p66shc, oxygen consumption is reduced 30% to 50% with a coincident increase in glycolysis and
decrease in NADH metabolism. In addition, Giorgio et al. [27] demonstrated that
p66shc utilizes reducing equivalents of the mitochondrial electron transfer chain
through the oxidation of cytochrome c to generate mitochondrial H2 O2, which functions as a signaling molecule for apoptosis.
7 Conclusion
The role of ROS in the process of carcinogenesis has evolved from damaging lipid,
proteins, and nucleic acids that in turn promote tumor development to regulators of
signal transduction events that control key facets of the malignant phenotype. Many
of the redox-sensitive signaling proteins drive the expression or activation of genes
and proteins that regulate tumor cell proliferation, survival, vascular recruitment,
and migration. The source of the ROS that regulate these signaling networks has in
many cases been assigned to the ubiquitous family of noninflammatory respiratory
burst oxidases. However, emerging data indicates that mitochondrial-derived ROS
also play a prominent role in regulating the tumorigenic phenotype. A number of
Respiratory Control of Redox Signaling and Cancer
43
signaling proteins like the dual lipid protein phosphatase PTEN and the transcription
factor Ets-1 are sensitive to perturbations in mitochondrial oxidant production and
are involved in regulating numerous aspects of the tumorigenic phenotype. H2 O2 dependent activation of Ets-1 leads to enhanced tumor angiogenesis, migration, and
invasion. Oxidative inactivation of PTEN leads to an increase in active AKT thereby
promoting tumor cell survival, proliferation, and angiogenesis. This implies that
alterations in mitochondrial function that enhance respiratory ROS production can
control signaling molecules that synergize and potentially drive the malignant phenotype. PTEN contains an active site cysteine that renders it redox-sensitive. Ets-1
contains several cysteines that are essential for its DNA binding, and it is likely that
oxidation would lead to its inactivation. The fact that oxidants promote Ets-1 activity
suggest some as yet undefined oxidant-sensitive signaling molecule may control the
redox-responsiveness of this transcription factor independent of its DNA-binding
activity.
It is clear that many facets of the tumorigenic phenotype are ROS responsive.
The angiogenic factor VEGF is controlled by oxidants through promoter activation,
stabilization of transcription factors (HIF-1␣), and activation of upstream signals.
Ets-1 activation also contributes to the early phases of blood vessel formation. Thus,
angiogenic signals appear to be highly sensitive to shifts in oxidant production and
may be highly responsive to shifts in respiratory oxidant production. Further characterization and identification of oxidant-responsive molecular targets that control
tumor progression and eradication of cancer stem cells will undoubtedly lead to the
development of novel antioxidant-based therapies for the treatment of malignancy.
References
1. Nemoto S, Takeda K, Yu ZX, Ferrans VJ, Finkel T. Role for mitochondrial oxidants as regulators of cellular metabolism. Mol Cell Biol 2000; 20(19):7311–7318.
2. Rhee SG, Chang TS, Bae YS, Lee SR, Kang SW. Cellular regulation by hydrogen peroxide. J
Am Soc Nephrol 2003; 14(8 Suppl 3):S211–S215.
3. Wood ZA, Poole LB, Karplus PA. Peroxiredoxin evolution and the regulation of hydrogen
peroxide signaling. Science 2003; 300(5619):650.
4. Poole LB. Bacterial defenses against oxidants: mechanistic features of cysteine-based peroxidases and their flavoprotein reductases. Arch Biochem Biophys 2005; 433(1):240–254.
5. Tonks NK. Protein tyrosine phosphatases: from genes, to function, to disease. Nat Rev Mol
Cell Biol 2006; 7(11):833–846.
6. Arnold RS, Shi J, Murad E, et al. Hydrogen peroxide mediates the cell growth and transformation caused by the mitogenic oxidase Nox1. Proc Natl Acad Sci USA 2001; 98(10):
5550–5555.
7. Tonks NK. Protein tyrosine phosphatases: from genes, to function, to disease. Nat Rev Mol
Cell Biol 2006; 7(11):833–846.
8. Floyd RA. Antioxidants, oxidative stress, and degenerative neurological disorders. Proc Soc
Exp Biol Med 1999; 222(3):236–245.
9. Menon SG, Goswami PC. A redox cycle within the cell cycle: ring in the old with the new.
Oncogene 2007; 26(8):1101–1109.
10. Adams JM. Ways of dying: multiple pathways to apoptosis 2. Genes Dev 2003; 17(20):
2481–2495.
44
P.M. Carrico et al.
11. Wong GHW, Goeddel DV. Induction of manganous superoxide dismutase by tumor necrosis
factor: possible protective mechanism. Science 1988; 242:941–944.
12. Dasgupta J, Subbaram S, Connor KM, et al. Manganese superoxide dismutase protects from
TNF-alpha-induced apoptosis by increasing the steady-state production of H2 O2 . Antioxid
Redox Signal 2006; 8(7–8):1295–1305.
13. Chiarugi P, Buricchi F. Protein tyrosine phosphorylation and reversible oxidation: two crosstalking posttranslation modifications. Antioxid Redox Signal 2007; 9(1):1–24.
14. Lambeth JD, Cheng G, Arnold RS, Edens WA. Novel homologs of gp91phox. Trends
Biochem Sci 2000; 25(10):459–461.
15. Nelson KK, Melendez JA. Mitochondrial redox control of matrix metalloproteinases. Free
Radic Biol Med 2004; 37(6):768–784.
16. Lincoln DW, Bove K. The transcription factor Ets-1 in breast cancer. Front Biosci 2005;
10:506–511.
17. Connor KM, Subbaram S, Regan KJ, et al. Mitochondrial H2 O2 regulates the angiogenic
phenotype via PTEN oxidation. J Biol Chem 2005; 280(17):16916–16924.
18. Brunelle JK, Bell EL, Quesada NM, et al. Oxygen sensing requires mitochondrial ROS but
not oxidative phosphorylation. Cell Metab 2005; 1(6):409–414.
19. Nelson KK, Ranganathan AC, Mansouri J, et al. Elevated sod2 activity augments matrix metalloproteinase expression: evidence for the involvement of endogenous hydrogen peroxide in
regulating metastasis. Clin Cancer Res 2003; 9(1):424–432.
20. Kim JH, Chu SC, Gramlich JL, et al. Activation of the PI3K/mTOR pathway by BCRABL contributes to increased production of reactive oxygen species. Blood 2005; 105(4):
1717–1723.
21. Brunet A, Bonni A, Zigmond MJ, et al. Akt promotes cell survival by phosphorylating and
inhibiting a Forkhead transcription factor. Cell 1999; 96(6):857–868.
22. Paik JH, Kollipara R, Chu G, et al. FoxOs are lineage-restricted redundant tumor suppressors
and regulate endothelial cell homeostasis. Cell 2007; 128(2):309–323.
23. Tothova Z, Kollipara R, Huntly BJ, et al. FoxOs are critical mediators of hematopoietic stem
cell resistance to physiologic oxidative stress. Cell 2007; 128(2):325–339.
24. Nemoto S, Finkel T. Redox regulation of forkhead proteins through a p66shc-dependent signaling pathway. Science 2002; 295(5564):2450–2452.
25. Sansone P, Storci G, Giovannini C, et al. p66Shc/Notch-3 interplay controls self-renewal and
hypoxia survival in human stem/progenitor cells of the mammary gland expanded in vitro as
mammospheres. Stem Cells 2007; 25(3):807–815.
26. Nemoto S, Combs CA, French S, et al. The mammalian longevity-associated gene product
p66shc regulates mitochondrial metabolism. J Biol Chem 2006; 281(15):10555–10560.
27. Giorgio M, Migliaccio E, Orsini F, et al. Electron transfer between cytochrome c and
p66Shc generates reactive oxygen species that trigger mitochondrial apoptosis. Cell 2005;
122(2):221–233.
Cellular Respiration and Dedifferentiation
Roberto Scatena, Patrizia Bottoni, and Bruno Giardina
Abstract Mitochondria are far more than the “powerhouse” of the cell as they
have classically been described. They are also the location where various catabolic
and anabolic processes, calcium fluxes, reactive oxygen and nitrogen species, and
numerous signal transduction pathways interact to maintain cell homeostasis and
to modulate cellular responses to diverse stimuli. Because of the mitochondrion’s
vital roles in cell physiology, this organelle squarely falls within the ambit of cancer pathophysiology. Indeed cancer-associated alterations in cell energy metabolism
related to mitochondrial dysfunction have been recognized for many years in the socalled Warburg effect. Furthermore, these organelles appear to play a fundamental
role in determining whether a cell will undergo cell death by apoptosis or necrosis;
and key metabolic enzymes of the tricarboxylic acid (TCA) cycle, the fundamental
oxidative pathway of the cell, may act as oncosuppressors. Recent findings indicate
that modulation of mitochondrial activities in general, and of its electron respiratory
chain in particular, could induce differentiation and/or death of cancer cells. Thus
it may be possible to induce anticancer activities by deliberate modulation of cellular respiration. Elucidation of the role of mitochondria in cancer cell dedifferentiation processes may provide a new definition of the dedifferentiation/differentiation
of cancer cells as well as important information about the pathophysiology
of so-called stem cells, with important diagnostic, prognostic, and therapeutic
consequences.
Keywords Cancer cell differentiation · Oxidative phosphorylation · NADH
dehydrogenase · Heat shock proteins · Reactive oxygen species · Warburg effect ·
Nitric oxide · Oncogenes · Oncosuppressors · Biomarkers
R. Scatena (B)
Dipartimento di Medicina di Laboratorio, Universit`a Cattolica del Sacro Cuore, 0168 Rome, Italy
e-mail: r.scatena@rm.unicatt.it
S.P. Apte, R. Sarangarajan (eds.), Cellular Respiration and Carcinogenesis,
DOI 10.1007/978-1-59745-435-3 4,
C Humana Press, a part of Springer Science+Business Media, LLC 2009
45
46
R. Scatena et al.
1 Introduction
Cancer can be defined as a genetic disease characterized by dysregulation of various cellular pathways that orchestrate cell proliferation, differentiation, and death
[1]. A cell becomes cancerous when a series of oncogenes and/or oncosuppressors become dysfunctional and hence induce a neoplastic phenotype in that cell. In
this dramatic clonal disorder, the cell acquires high proliferation kinetics and loses
cell-cell contact inhibition. These properties are progressively achieved by a series
of genetic “hits,” including mutations and or other epigenetic modifications [1].
The final result is often defined as dedifferentiation of a neoplastic cell that morphologically becomes anaplastic. The grade of dedifferentiation (anaplasia) differs
among cancer cell specimens and is often correlated with the malignancy of the
neoplasia.
In the dramatic perturbation of cell homeostasis that accompanies carcinogenesis, mitochondria are involved in various relevant, interconnected, but also indirect
pathophysiologic roles. In this context, the significance of mitochondrial respiration
has received inadequate attention. In spite of an abundance of data that link mitochondria to cancer, there have been few studies that have investigated the interrelationships between cellular respiration and cellular dedifferentiation. Here we briefly
summarize some molecular mechanisms, which will be extensively considered in
other chapters, that link cancer cell dedifferentiation to mitochondrial oxidative
metabolism to provide an introduction to concepts that facilitate the understanding
of these aspects of oncology:
1. Peculiar modification of cell metabolism (Warburg effect). Cancer cells generally produce ATP mainly through so-called aerobic glycolysis, a metabolic shift
characterized by high glucose uptake and increased production of lactate. Originally, this altered metabolic state was considered to be the result of adaptation
(via HIF and/or AKT) to the new potentially anoxic environment of a neoplastic
lesion [2]. However it has become evident that the aerobic glycolysis of cancer
cells is an epiphenomenon that results from a more complex metabolic rearrangement in which not only glycolysis but also Krebs cycle, beta-oxidation,
and anabolic metabolism in general are altered to support the new primary function of the cell (e.g., uncontrolled proliferation) by providing not only energy
but also the building blocks for the synthesis of nucleic acids, peptides, and
lipids [3–5].
2. Distinct properties of cancer cell mitochondria. Emerging data suggest that mitochondria have an active pathogenic role in the induction and maintenance of the
Warburg effect. This role is emphasized by some typical metabolic peculiarities of mitochondria that enable cancer cells to be functionally distinguished
from normal cells. Cancer cells exhibit reduced mitochondrial respiration (both
in terms of ATP turnover and proton leakage), increased NADH/NAD+ ratios,
and a pronounced increase in membrane potential [6]. Other morphologic
(number, size, shape) and structural differences (phospholipids, cholesterol, and
Cellular Respiration and Dedifferentiation
3.
4.
5.
6.
47
membrane proteins) have been described, but their pathogenic roles are not
understood [7].
Mitochondrial DNA (mtDNA) mutations. Mutations of mtDNA have been
linked to some tumors. In particular, intragenic deletions, missense and chainterminating point mutations, and alterations of homopolymeric sequences have
been described in various sporadic tumors. Intriguingly, mutations have been
described in the hypervariable regions of the mitochondrial D-loop, the section
of DNA that controls mtDNA transcription and replication. Mutations have also
been described in all tRNAs, in rRNAs, and in all 13 of the mtDNA-encoded
subunits of the electron respiratory chain [8]. However, Salas et al. [9] recently
suggested that most of the reported mtDNA mutations in tumors may be due
to laboratory-induced artifacts. Others contend that mtDNA mutations play a
real pathogenetic role in cancer [10]. Hence there is ongoing debate about the
role of mtDNA mutations and the consequent respiratory dysfunction in cancer
cells.
Nuclear DNA mutations and mitochondrial proteins. Succinate dehydrogenase
(SDH) and fumarate hydratase (FH), mitochondrial enzymes of the tricarboxylic
acid (TCA) cycle, can act as classic tumor suppressors, and mutations of the
genes that encode them have been associated with various benign and malignant
tumors. These observations have drawn attention to the potential pathogenetic
role of the mitochondrion, especially of mitochondrial oxidative metabolism,
in cancer [11, 12]. Given the role of SDH (complex II) in the electron respiratory chain, such findings suggest that there may be critical interrelationships
between mitochondrial energy metabolism and timely surveillance of the differentiation/dedifferentiation state of cells.
Apoptosis. Mitochondria have a well-defined role in caspase-mediated induction of the intrinsic pathway of apoptosis initiated by the release of cytochrome
c and second mitochondrial activator of caspases (SMAC) from the mitochondrial intermembrane space in response to a variety of noxious stimuli, including
DNA damage, loss of adherence to the extracellular matrix (ECM), oncogeneinduced proliferation, and growth factor deprivation. Perturbation in the release
of apoptosis-inducing proteins regulated by proapoptotic and antiapoptotic members of the Bcl-2 family (e.g., Bcl-2, Bcl-XL, and Mcl-1) seems to influence
not only etiopathogenesis but also cancer prognosis and therapeutic efficacy.
Accordingly, derangements of the permeability transition pore (PTP) and electrochemical gradient, which strictly depend on mitochondrial respiration, have
been implicated in these phenomena [13].
Free radicals (reactive oxygen species [ROS] and reactive nitrogen species
[RNS]). Mitochondria are the main source of free radicals in cells. Thus mitochondrial dysfunction, especially at the level of the electron respiratory chain,
can result in increased production of ROS and RNS. ROS are considered classic
carcinogens because they can directly induce mutagenesis and thereby promote
tumorigenesis and cancer progression [14]. However, the role of ROS and RNS
in cancer is complex. In fact, the growth-promoting effects of ROS have been
also related to activation of redox-responsive signaling cascades (MAPK, PKC,
48
R. Scatena et al.
AP-1, NF-␬B). Paradoxically, ROS can also act as tumor suppressors, activating via a p16ink4a -Rb pathway an irreversible cellular process of senescence that
establishes stable G1 phase cell-cycle arrest [15].
7. Heat shock proteins (HSPs). HSPs represent a ubiquitous and evolutionarily
conserved family of proteins (classified according to their molecular weight)
that were originally discovered because they can be induced by heat stress.
Their production confers a peculiar cellular resistance to further stresses (heat or
chemical). Initially, these proteins were considered to be chaperones—molecular
guides acting to ensure proper protein folding. Other functions have since been
discovered, including roles in gene expression regulation, DNA replication, signal transduction, cell differentiation, metastasis, apoptosis, senescence, immortalization, and, intriguingly, multidrug resistance development. However, the
precise pathophysiologic roles of HSPs in general, and of mitochondrial Heat
Shock Proteins (mitHSPs) in particular, are not well understood. Studies employing several methodological approaches (immunochemical, immunohistochemical, proteomic) have shown that HSP expression gradually increases from normal
to dysplastic and then to neoplastic tissues, such that HSP overexpression correlates with cancer evolution. Hence some of these proteins (i.e., HSP60, HSP10,
HSP70, PHB1, and PHB2) have been implicated in cancer diagnosis, prognosis,
and therapy [16]. Overexpression of HSP70 (also known as 75-kDa glucose regulated protein, HSP70-9, HSP70-9B, mortalin 2) may be particularly important
in cancer cell proliferation and dedifferentiation; its role may go beyond generic
cytoprotection with respect to apoptosis and involve a more specific role in the
pathogenesis of the peculiar cellular metabolic shift that characterizes the cancer
cell phenotype [17].
8. Hexokinase II. The cancer cell glycolytic phenotype is supported by overexpression of the hexokinase II isoenzyme, which is linked to the outer mitochondrial
membrane via the porin-like protein voltage-dependent anion channel (PPVAC).
Hexokinase II-PPVAC binding may have significant functional implications in
cancer because it facilitates glucose phosphorylation and thus glucose utilization
for energetic and biosynthetic purposes [18].
9. Calcium. Because of the pivotal roles mediated by calcium ions in signal
transduction pathways, calcium has been considered to be a potential physiopathogenetic marker in several chronic and acute cellular diseases. However, research findings have not confirmed the originally grand expectations
for calcium in disease processes, at least in oncology. When Scorrano et al.
[19] examined the interrelationships between calcium, mitochondrial respiration, and dedifferentiation, they found only an indirect role of Bax/Bak via
regulation of calcium release from the endoplasmic reticulum. Interestingly,
however, chemotherapeutic-resistant tumor cells sequester cytosolic calcium
(Cacyt ) more efficiently than do normal cells; they show reduced release of calcium ions from intracellular storage sites upon apoptosis induction [20]. It has
been proposed that transition to more invasive cellular behavior may be related
to impaired release of calcium from the endoplasmic reticulum of various kinds
of cancer cells. On the contrary, Amuthan et al. [21] showed in tumorigenic
Cellular Respiration and Dedifferentiation
49
human pulmonary carcinoma A549 cells that mitochondrial stress induced by
damaging mtDNA (with ethidium bromide treatment) activates a mitochondria
to nucleus signaling pathway mediated by increased Cacyt concentration and that
activation of this pathway induces cell dedifferentiation associated with phenotypic changes and cell invasion.
2 Cellular Respiration and Dedifferentiation
Although the connection between mitochondrial respiration and cancer cell dedifferentiation (i.e., resistance to apoptosis, cell proliferation, metabolic shift, cell
invasion, genetic instability) is fundamental to cancer pathogenesis, the precise
molecular interrelationships and regulatory processes are far from clarified. This
lack of understanding is related, at least in part, to the traditional view that the socalled Warburg effect was essentially a reflection of mitochondria being damaged
and dysfunctional organelles in cancer cells. Recent studies are reexamining mitochondrial function, especially cellular respiration, in cancer. However, a consensus
among the findings has yet to be achieved [6, 21–25].
Herein some important findings linking cellular respiration to cell dedifferentiation in cancer are provided. It is worth noting that this putative link may have
crucial pharmacologic and clinical implications. The work of Amuthan et al. [21] is
particularly relevant in this regard because the findings provide mechanistic insight
into how cancer progression and tumor invasion may be directly related to chemical damage of mtDNA that results in mitochondrial respiratory chain dysfunction,
loss of mitochondrial membrane potential, and reduction of ATP synthesis. This
mitochondrial derangement also causes calcium release into the cytoplasm, which
activates a cell dedifferentiation program. Amuthan et al. [21] showed in C2C12
myoblasts and in human pulmonary carcinoma A549 cells that thus-induced dedifferentiation programs can transform cells into a more aggressive phenotype with the
following properties:
• induction of hexokinase, the enzyme that primes glucose by consuming ATP;
• induction of phosphoenolpyruvate carboxykinase (PEPCK), which regulates gluconeogenesis, glyceroneogenesis, and many other anaplerotic reactions;
• induction of tumor invasion markers (i.e., cathepsin L, TGF-␤1, ERK1, ERK2,
calcineurin) and antiapoptotic proteins (Bcl2 and Bcl-XL );
• reduction of levels of the proapoptotic proteins Bid and Bax;
• reduction of cell growth and occurrence of morphologic changes.
In summary, the findings of Amuthan et al. [21] demonstrate that, at least in this
set of experimental conditions, a generic mitochondrial stress caused by mtDNA
depletion deranges cellular respiration and alters patterns of nuclear gene expression
and thereby induces tumorigenic properties.
50
R. Scatena et al.
In general, it is well-known that mtDNA depletion can also render cells more
sensitive to apoptotic stimuli. Amuthan et al. [21] explain that the outcome of the
stress (cell dedifferentiation vs. apoptosis) could be related to the relative resistance
to apoptosis and/or to the capacity of the adopted cellular models to buffer efficiently
the secondary metabolic stresses. Furthermore, evidence obtained from this combination of methodological approaches demonstrated the existence of mitochondriato-nucleus stress signaling, through which alterations in cellular respiration have
the ability to change nuclear gene expression and selectively promote the growth of
more aggressive clones.
These results compel us to consider other research demonstrating that mitochondria can mediate the proliferative or inhibitory actions of different physiologic or
pharmacologic molecules that influence cellular respiration. Interestingly, it has
been reported recently that some “nongenomic” effects of estrogens, that is; effects
mediated by mitochondria rather than estrogen receptors, seem to modulate the
expression of nuclear cell cycle genes and human breast tumor growth. Estrogens
cause increases in the transcription of various complex subunits of the electron respiratory chain in mitochondria. For example, estradiol treatment (0.5 nM for 6 days)
is associated with a 16-fold increase in cytochrome oxidase II (COII) mRNA in rat
pituitary tumor cells. Similarly, ethinyl estradiol treatment (0.5 to 10 ␮M for 40
hours) caused a two- to threefold increase in cytochrome oxidase I (COI), COII, and
NADH dehydrogenase subunit I. Estrogens seem to negatively affect mitochondria
at the protein level as well. Several studies have shown that estrogens can inhibit
mitochondrial respiratory complexes I, II, III, IV, and V. This perturbation of the
electron respiratory chain can induce mitochondrial ROS production, which could
promote the generation and progression of some cancers via oxidative stress and/or
ROS-related signaling pathways [23]. Given the suggested role of estrogens in some
cancers (i.e., breast cancer) and the discussed pathogenetic mechanisms at the basis
of oncopromotion and oncoprogression activity, further evidence should be gathered
to determine the relevance of these findings.
The ability of lipophilic molecules to derange cellular respiration and alter cell
differentiation/dedifferentiation processes is a neglected pathophysiologic facet of
the mitochondrial respiratory chain. On the basis of the cytopathology of liver disease associated with use of fibrates (lipid-reducing drugs) and some experimental
data indicating that fibrate binding to human hemoglobin can functionally modify the hemoprotein, we postulated that fibrates, and their thiazolidinedione derivatives, could interact with many hydrophobic protein domains and especially with
complex I subunits of the mitochondrial electron respiratory chain. We found that
this molecular interaction is rapid (<1 hour from administration) and occurs in
numerous cell types in varying degrees. The following human cell lines showed
derangement of varying degrees of complex I and cell differentiation: K-562
human erythroleukemia, HL-60 human myeloid leukemia, U-937 human monocytic
leukemia, TE-671 human rhabdomyosarcoma, HepG2 human hepatocarcinoma, and
SK-N-BE[2] human neuroblastoma [22, 24, 25].
In fact, these molecules, by their lipophilic properties, diffuse freely into lipid
bilayers in general and in mitochondria in particular also because of interplay
Cellular Respiration and Dedifferentiation
51
between their logD and mitochondrial electrochemical gradient, altering in such
a way the function of proteins embedded in membranes. One of the larger protein
components of mitochondria membranes is complex I (NADH-ubiquinone oxidoreductase, subunits >30, MW 850 kDa), which because of these features may act as
the main lipophilic molecular sink in mitochondria [26].
The rapid iatrogenic respiratory derangement caused by introduction of these
lipophilic molecules can induce metabolic and ROS-mediated stress that provokes
a differentiation process, also mitochondrial HSP-mediated, in human tumor cell
lines. In particular, the derangement of mitochondrial complex I may induce ROS
production and may impair NADH oxidation resulting in an energetic shortage in
cancer cells. Importantly, considering that complex I is only partially deranged, the
oxidative and energetic stress are in general not lethal. In this condition, cell growth
is hampered and phenotypic differentiated features may reappear. This implies that
there is modification of the actual differentiation/dedifferentiation definition and
thus suggests that a redefinition of the role of some oncogenes and oncosuppressors is appropriate. Data confirming that fibrates and thiazolidinediones inhibit, via
nongenomic activity, complex I–impairing NADH oxidation at the mitochondrial
level can explain some of the pharmacologic activities of this class of drugs (i.e.,
antihyperlipidemia, insulin sensitization). The findings also may account for some
adverse side effects of fibrates and thiazolidinediones, including myocardiotoxicity,
hepatotoxicity, and, most relevant to this discussion, peroxisome proliferation, carcinogenicity, and cancer differentiating properties, considered typical peroxisome
proliferator activated receptor (PPAR)-␣–mediated effects in rodents [27].
The differentiating activity of fibrates and thiazolidinediones on human tumor
cell lines has been shown to be tightly correlated with the level of mitochondrial
complex I derangement. This differentiating activity is independent of the expression level of PPARs, as well as the embryonic origin of the tissue. Stress-related differentiation in human tumor cell lines, including some human leukemia cell lines,
is a well-known occurrence [28].
Fibrate/thiazolidinedione-induced differentiation is unusual; it is linked to
the metabolic status of differentiated cells that shift toward a more glycolytic
metabolism (paradoxical Warburg effect) and thus showing that original neoplastic
phenotype yet used mitochondria. Moreover, metabolic parameters recorded during drug-induced cell differentiation (residual activity of the mitochondrial electron
respiratory chain components glycerol 3-phosphate dehydrogenase and succinate
dehydrogenase; levels of glycolytic and nonglycolytic acetate, pyruvate, and alanine) suggest that the so-called Warburg effect is really an epiphenomenon of a more
complex metabolic rearrangement of the neoplastic cell. This cancer-associated
metabolic rearrangement is induced by the cancer’s rigid genetic selection program,
characterized by a high level of cell proliferation and diffusion, which is normally
present only during embryonic and fetal development.
Our understanding of the functional connections between alterations in cell
metabolism and mitochondrial physiology in cancer is under profound revision. And
all mitochondrial physiology is strongly affected by electron respiratory chain activities. The recent results of Fantin et al. [6] are consistent with the view that tumor
52
R. Scatena et al.
progression is strongly associated with the so-called Warburg effect and mitochondrial metabolism. These authors showed that attenuation of lactate dehydrogenase A
(LDH-A) expression by LDH-A short hairpin RNAs in tumor cell lines reactivates
mitochondrial respiration, in terms of proton leakage and ATP turnover, and restores
the normally depolarized state of mitochondrial membranes. In this situation, cancer
cells can no longer oxidize NADH via LDH but must rely on the mitochondrial respiratory chain. Interestingly, this abruptly induced change in metabolism inhibited
cell growth. The authors concluded that these results demonstrate a fundamental
role of aerobic glycolysis in general and LDH in particular in tumor maintenance
and progression and further show that cancer-associated slowdown of mitochondrial
respiration is not due to structural damage of the organelles but rather results from
altered metabolic signals. NADH/NAD+ ratio and modulation of related mitochondrial shuttle systems are likely involved in these signals.
The fundamental role of cellular respiration in cancer cell dedifferentiation is further emphasized by a recent study that examined energy metabolism in breast cancer
brain metastasis, which undoubtedly represents a clear in vivo model of bioenergetic adaptation of dedifferentiated cells [29]. Interestingly, these metastatic cells
showed increased expression of enzymes involved in glycolysis, in accordance with
the Warburg effect, but also showed increased expression of enzymes involved in
the TCA cycle, pentose phosphate pathways, beta-oxidation, and oxidative phosphorylation. Intriguingly, the authors also reported induction of enzymatic pathways
related to glutathione synthesis, which together with activation of the pentose phosphate pathway could indicate the presence of antioxidant defense mechanisms to
minimize production of ROS derived from enhanced oxidative metabolism [29].
3 Conclusion
Emerging evidence, despite conflicts likely due to experimental differences, is shedding light on the interrelationships between cellular respiration and cancer cell dedifferentiation, in particular:
• Confirms a more active role of mitochondrial respiration to maintain the cellular
differentiated phenotype and at the same time underline the pathogenetic significance of perturbations of mitochondrial electron respiratory chain in promotion
and progression of cancer.
• Stresses the importance of the shift of all cellular metabolism to ensure the new
energetic and structural needs of neoplastic cell.
• Justifies some debated metabolic and functional evidence peculiar to cancer cell
(i.e., induction of hexokinase II and its binding to ANT).
• Enhances the role of NADH and NADH/NAD+ ratio in modulation of a signal
transduction pathway that controls differentiation/dedifferentiation status.
• Emphasizes the interplay of complex I and complex II to adapt cell metabolism
to different cellular physiopathologic conditions.
Cellular Respiration and Dedifferentiation
53
• Last but not least, addresses the new research, which explores the interrelationships between cellular respiration and cell dedifferentiation, toward the regulatory function of complex I, complex II, and above all pushes to investigate
deeper, in terms of cancer cell metabolism adaptions, the other pathways of electron transport (i.e., alpha glycerophosphate and fatty acyl CoA).
In conclusion, it is clear that there exists a significant link between cellular respiration and cell differentiation/dedifferentiation. Elucidating the molecular mechanisms at the basis of this link may have critical clinical implications in terms of
cancer diagnosis, prognosis, and, most importantly, therapy. Disentangling the relationship between cancer cell oxidative metabolism and cellular differentiation state
is fundamental not only for oncology but also for properly understanding and using
all potential clinical applications related to the physiopathology of stem cells.
References
1. Bale AE, Brown SJ. Etiology of cancer: cancer genetics. In: De Vita VT, Hellman S,
Rosenbreg SA, eds. Cancer—Principles & practice of oncology, 6th ed. Philadelphia:
Lippincot Williams & Wilkins, 2001:170–183.
2. Gatenby RA, Gillies RJ. Why do cancers have high aerobic glycolysis? Nat Rev Cancer 2004;
4:891–899.
3. Moreno-Sanchez R, Rodriguez-Enriquez S, Marin-Hernandez A, Saavedra E. Energy
metabolism in tumor cells. FEBS J 2007; 274:1393–1418.
4. Kim JW, Dang CV. Cancer’s molecular sweet tooth and the Warburg effect. Cancer Res 2007;
66:8927–8930.
5. Garber K. Energy deregulation: licensing tumor to grow. Science 2006; 312:1158–1159.
6. Fantin VR, St-Pierre J, Leder P. Attenuation of LDH-A expression uncovers a link between
glycolysis, mitochondrial physiology, and tumor maintenance. Cancer Cell. 2006; 9:425–434.
7. Modica-Napolitano JS, Singh KK. Mitochondrial dysfunction in cancer. Mitochondrion 2004;
4:755–762.
8. Brandon M, Baldi P, Wallace DC. Mitochondrial mutations in cancer. Oncogene 2006 7;
25:4647–4662.
9. Salas A, Yao YG, Macaulay V, Vega A, Carracedo A, Bandelt HJ. A critical reassessment of
the role of mitochondria in tumorigenesis. PLoS Med 2005; 2(11):e296.
10. Herrnstadt C, Preston G, Howell N. Errors, phantoms and otherwise, in human mtDNA
sequences. Am J Hum Genet 2003; 72:1585–1586.
11. King A, Selak MA, Gottlieb E. Succinate dehydrogenase and fumarate hydratase: linking
mitochondrial dysfunction and cancer. Oncogene 2006; 25:4675–4682.
12. Gottlieb E, Tomlinson IP. Mitochondrial tumour suppressors: a genetic and biochemical
update. Nat Rev Cancer 2005; 5:857–866.
13. Cheng WC, Berman SB, Ivanovska I, et al. Mitochondrial factors with dual roles in death and
survival. Oncogene 2006; 25:4697–4705.
14. Droge W. Free radicals in the physiological control of cell function. Physiol Rev 2002; 82:
47–95.
15. Takahashi A, Ohtani N, Yamakoshi K, et al. Mitogenic signalling and the p16INK4a-Rb pathway cooperate to enforce irreversible cellular senescence. Nat Cell Biol 2006; 8:1291–1297.
16. Calderwood SK, Khaleque MA, Sawyer DB, Ciocca DR. Heat shock proteins in cancer: chaperones of tumorigenesis. Trends Biochem Sci 2006; 31:164–172.
54
R. Scatena et al.
17. Czarnecka AM, Campanella C, Zummo G, Cappello F. Mitochondrial chaperones in cancer:
from molecular biology to clinical diagnostics. Cancer Biol Ther 2006; 5:714–720
18. Mathupala SP, Ko YH, Pedersen PL. Hexokinase II: cancer s double-edged sword acting as
both facilitator and gatekeeper of malignancy when bound to mitochondria. Oncogene 2006;
25(34):4777–4786.
19. Scorrano L, Oakes SA, Opferman JT, et al BAX and BAK regulation of endoplasmic reticulum
Ca2+: a control point for apoptosis. Science 2003; 300:135–139.
20. Chandra D, Liu JW, Tang DG. Early mitochondrial activation and cytochrome c up-regulation
during apoptosis. J Biol Chem 2002; 277:50842–50854.
21. Amuthan G, Biswas G, Ananadatheerthavarada HK, Vijayasarathy C, Shephard HM,
Avadhani NG. Mitochondrial stress-induced calcium signaling, phenotypic changes and invasive behavior in human lung carcinoma A549 cells. Oncogene 2002; 21:7839–7849.
22. Scatena R, Bottoni P, Vincenzoni F, et al. Bezafibrate induces a mitochondrial derangement in
human cell lines: a PPAR-independent mechanism for a peroxisome proliferator. Chem Res
Toxicol 2003; 16:1440–1447.
23. Felty Q, Roy D. Estrogen, mitochondria, and growth of cancer and non-cancer cells.
J Carcinog 2005; 4:1–18
24. Bottoni P, Giardina B, Martorana GE, et al. A two-dimensional electrophoresis preliminary
approach to human hepatocarcinoma differentiation induced by PPAR-agonists. J Cell Mol
Med 2005; 9:462–467.
25. Scatena R, Nocca G, Sole PD, et al. Bezafibrate as differentiating factor of human myeloid
leukemia cells. Cell Death Differ 1999; 6:781–787.
26. Scatena R, Bottoni P, Martorana GE, et al. Mitochondria, ciglitazone and liver: a neglected
interaction in biochemical pharmacology. Eur J Pharmacol 2007; 567:50–58.
27. Michalik L, Desvergne B, Wahli W. Peroxisome proliferator-activated receptor and cancers:
complex stories. Nat Rev Cancer 2004; 4:61–70.
28. Richards FM, Watson A, Hickman JA. Investigation of the effects of heat shock and agents
which induce a heat shock response on the induction of differentiation of HL-60 cells. Cancer
Res 1988; 48:6715–20.
29. Chen EI, Hewel J, Krueger JS, et al. Adaptation of energy metabolism in breast cancer brain
metastases. Cancer Res 2007; 67:1472–1486.
Cellular Adaptations to Oxidative
Phosphorylation Defects in Cancer
Sarika Srivastava and Carlos T. Moraes
Abstract Mitochondrial DNA (mtDNA) somatic mutations or mutations in nuclear
genes encoding mitochondrial proteins important for the assembly, activity, or maintenance of the individual oxidative phosphorylation (OXPHOS) complexes have
been observed in tumors. Although the functional consequence of such mutations
is unclear at the moment, retrograde signaling in response to OXPHOS defects
can activate various nuclear genes and signaling pathways that alter mitochondrial function, tumor invasion, metastasis, redox-sensitive pathways, programmed
cell death pathways, calcium signaling pathways, and cellular pathways leading
to global changes in cellular morphology and architecture. In addition, we have
found that some cancer cell lines harboring deleterious mtDNA mutations upregulate the expression of members of the peroxisome-proliferator activated ␥ coactivator 1 family of coactivators, probably to sustain the necessary ATP production for
cell proliferation. In this chapter, we describe such cellular adaptations and changes
in response to OXPHOS defects that are associated with a variety of cancer cell
types.
Keywords mtDNA · Retrograde signaling · OXPHOS · Cancer · Mitochondria
1 Mammalian Oxidative Phosphorylation
Mammalian mitochondrial oxidative phosphorylation (OXPHOS) occurs in five
multimeric enzyme complexes (complexes I, II, III, IV, and V) using two electron
transport carriers (ubiquinone or coenzyme Q10 and cytochrome c). Electrons from
reduced equivalents (NADH and FADH2 ) are transported along these complexes via
ubiquinone and cytochrome c to molecular oxygen, producing water. At the same
time, protons are pumped across the mitochondrial inner membrane (from the matrix
to intermembrane space) by enzyme complexes I, III, and IV thereby generating a
C.T. Moraes (B)
University of Miami Miller School of Medicine, Miami, FL 33136
e-mail: cmoraes@med.miami.edu
S.P. Apte, R. Sarangarajan (eds.), Cellular Respiration and Carcinogenesis,
DOI 10.1007/978-1-59745-435-3 5,
C Humana Press, a part of Springer Science+Business Media, LLC 2009
55
56
S. Srivastava, C.T. Moraes
Intermembrane space
H+
H+
H+
eO2
e-
e-
e-
e-
H2 O
e-
Cyt c
CoQ10
ATP
ADP
H+
Complex I
Complex II
Complex III
Complex IV
Complex V
Matrix
Fig. 1 Schematic representation of mammalian mitochondrial OXPHOS. Mammalian electron
transport chain (ETC) consists of four multimeric complexes (complexes I, II, III, and IV). Electrons from the reduced equivalents, NADH and FADH2 , enter the ETC and reduce complex I and
complex II, respectively. Coenzyme Q10 or ubiquinone is an electron carrier in the inner mitochondrial membrane that accepts an electron from either complex I or complex II and donates it
to complex III. Cytochrome c, another electron carrier in the mitochondrial intermembrane space,
accepts an electron from complex III and donates it to complex IV. Complex IV donates electrons
to molecular O2 , which results in the formation of H2 O. During the electron flow, complexes I, III,
and IV pump the protons from the matrix toward the intermembrane space thereby generating an
electrochemical gradient across the mitochondrial inner membrane. The energy in the electrochemical gradient is harnessed by complex V to generate ATP from ADP (oxidative phosphorylation)
proton gradient or membrane potential (⌬⌿m). The energy in this gradient is harnessed by complex V (ATP synthase) to synthesize ATP from ADP, a phenomenon
termed OXPHOS, during which protons flow back from the mitochondrial intermembrane space to the mitochondrial matrix (Fig. 1).
Mitochondria are also important for cytosolic Ca+2 buffering and are the predominant source of ROS production in most cell types. Superoxide anions are generated
at the level of complex I and complex III in the inner mitochondrial membrane.
2 OXPHOS Defects in Cancer
OXPHOS defects in cancer cells were first reported more than 50 years ago by
Warburg who stated that “cancer cells are impaired in respiratory chain function
and are very glycolytic” [1]. Studies on various types of cancer cells have found
Cellular Adaptations to OXPHOS Defects in Cancer
57
this hypothesis to be correct in several but not all tumor cells. Several independent
groups have also reported that many cancer cells possess a full complement of respiratory chain enzymes and can couple electron transport chain to ATP production
[2]. Isolated tumor mitochondria have also been found to respire on a variety of substrates. Specific changes that occur in the energy metabolism of tumor cells have not
yet been established.
In the past 10 years, studies on mitochondrial DNA (mtDNA) mutations that
cause OXPHOS defects and their potential role in cancer have regained momentum.
Mitochondrial DNA mutations have been increasingly identified in various types of
cancer cells [3, 4]. Polyak et al. were first to report a detailed analysis of somatic
mtDNA mutations in human colorectal cancer cells by sequencing their complete
mitochondrial genome [3] (somatic mtDNA mutations are those present in the tumor
tissue but absent from the adjacent normal tissue). After this initial finding, mtDNA
from various tumor cell types was sequenced by different groups, and mtDNA mutations have now been reported in esophageal, ovarian, thyroid, head, neck, lung, bladder, and renal cancer cells [4–6]. The majority of the mtDNA mutations reported in
different tumor cell types are transitions (G-to-A or T-to-C), a feature that is characteristic of reactive oxygen species (ROS)-derived mutations. Further, most of the
somatic mtDNA mutations have been reported to be homoplasmic suggesting their
dominance at intra- and intercellular levels.
2.1 Functional Significance of OXPHOS Defects in Cancer
Although in the past decade, the list of mtDNA mutations affecting OXPHOS in
cancer cells has increased steadily, the functional relevance of these mutations in
tumor promotion or formation process are yet obscure. Because the majority of the
somatic mtDNA mutations found in tumor cells have been reported to be homoplasmic, one of the existing hypotheses is that the homoplasmic levels of these mutations
were probably achieved through some kind of replication or growth advantage that
the mutant mtDNA harbored over the wild type in the tumor progenitor cells. It has
also been suggested that mtDNA mutations may cause a moderate increase in ROS
production that in turn stimulates cell growth [3, 4]. Low levels of ROS have been
shown to stimulate mitosis in various cell types [7, 8]. In contrast with this hypothesis, Coller and co-workershave demonstrated that there is a sufficient opportunity
for a tumor progenitor cell to achieve homoplasmy through unbiased mtDNA replication and segregation during cell division [9]. Coller et al.constructed a theoretical
model based on computer simulation studies and showed that there is a good correlation between the reported frequency of homoplasmic mutations in tumor cells
and the predicted frequency of homoplasmy in tumor progenitor cells that can be
attained in the absence of any selection [9]. The authors suggested that mtDNA
mutations such as silent substitutions of amino acid coding regions or alterations in
the length of polynucleotide tracts that are polymorphic in healthy individuals are
the ones that are unlikely to confer any selective advantage to the mtDNA or the
58
S. Srivastava, C.T. Moraes
host cell. However, the authors did not exclude the possibility that certain mtDNA
point mutations that cause functional alterations (e.g., mutations affecting the conserved amino acid regions in the coding region of a protein) might be segregated
in a nonrandom manner and may provide a selective growth advantage to the host
cell [9].
mtDNA mutations leading to OXPHOS dysfunction have also been shown to
increase tumorigenicity in certain tumor cell types. Mutations in the mtDNA COXI
gene were shown to increase tumorigenicity in prostate cancer patients [10]. Petros
et al. reported that ∼11% to 12% of prostate cancer patients harbor pathogenic
COXI mtDNA mutations that alter the conserved amino acid regions in the protein and suggested that these mutations may play a role in the etiology of prostate
cancer [10]. mtDNA ATP6 gene mutation (T8993G) has also been shown to increase
tumorigenicity by two independent groups. The first study was from Petros and coworkers who showed that PC3 prostate cancer cybrid cells harboring T8993G ATP6
mutant mtDNA when introduced in nude mice formed tumors that were ∼7 times
larger than cells harboring the wild-type mtDNA suggesting a role of mtDNA mutation in tumor promotion process [10]. They also showed that tumor cells harboring
ATP6 mutant mtDNA generated significantly higher levels of ROS than did those
harboring wild-type mtDNA, further suggesting that high levels of ROS promoted
tumor cell growth. The second study was from Shidara and co-workers. They constructed HeLa transmitochondrial cybrids harboring homoplasmic T8993G ATP6
mtDNA mutation and showed that these cybrids grew much faster in culture as
well as in nude mice forming tumors that were larger in size compared with the
cybrids harboring the wild-type mtDNA suggesting that the mtDNA ATP6 mutation
in these cells conferred a selective growth advantage [11]. Further, transfection of
a wild-type nuclear version of the mitochondrial ATP6 gene in these HeLa cybrids
harboring homoplasmic ATP6 mutant mtDNA followed by their transplantation in
nude mice was shown to slow down the tumor cell growth. Conversely, the expression of a nuclear version of the mutant ATP6 gene in the wild-type cybrids was
found to accelerate tumor cell growth thereby suggesting that the growth advantage
in these tumor cells depended on the mitochondrial ATP6 function [11].
OXPHOS dysfunction caused by partial mtDNA depletion (genetic stress) or
treatment with mitochondrial-specific inhibitors (metabolic stress) has been shown
to induce cell invasion and tumor progression in otherwise noninvasive cells [12,
13]. A correlation between OXPHOS impairment and tumor aggression has also
been reported in renal carcinomas [14]. Simonnet et al. studied the mitochondrial
respiratory chain enzyme content in three different renal tumors and found that the
OXPHOS impairment increased from the less aggressive to the most aggressive
types of renal carcinomas [14]. They also found that renal oncocytomas (benign
tumors) are exclusively deficient in mitochondrial complex I activity, whereas the
activity of all other OXPHOS complexes are increased in these tumors [15]. These
tumors further showed dense mitochondrial proliferation suggesting an increase
in mitochondrial biogenesis as an attempt to compensate for the potential loss of
OXPHOS function. Complex I deficiency was also found in normal cells adjacent to
the tumor tissue leading the authors to suggest that the complex I deficiency could be
Cellular Adaptations to OXPHOS Defects in Cancer
59
an early event in the formation of renal oncocytomas [15]. Similarly, OXPHOS dysfunction and mitochondrial proliferation have also been reported in thyroid oncocytomas [16].
The current studies on potential effects of mtDNA mutations and OXPHOS
dysfunction in cancer cells therefore suggest their role in tumor promotion and/or
progression. It is believed that nuclear DNA mutations and not mtDNA mutations control the tumor initiation process. Akimoto and co-workers have tested this
hypothesis [17]. They showed that cells carrying nuclear DNA from tumor cells
and mtDNA from normal cells form tumors in nude mice whereas those carrying
nuclear DNA from normal cells and mtDNA from tumor cells do not form tumors
[17]. However, mutations in the mitochondrial enzymes that are encoded by the
nuclear genome have been associated with certain types of cancer. Germ-line mutations in the tricarboxylic acid (TCA) cycle enzyme fumarate hydratase have been
shown to be associated with uterine fibroids, skin leiomyomata, and papillary renal
cell cancer [18], and mutations in the succinate dehydrogenase (SDH) subunits B,
C, and D have been shown to be associated with hereditary paragangliomas and
pheochromocytomas [19–21]. The potential role of these mutations in the tumor
formation process is yet unclear.
3 Cellular Adaptations to OXPHOS Defects in Cancer
The cross-talk between mitochondria and nucleus plays an important role in maintaining mitochondrial function and integrity. Mitochondrial dysfunction affects the
mitochondrial-nuclear cross-talk leading to altered signaling cascades (Fig. 2). The
nuclear-mitochondrial stress signaling, also called retrograde signaling, is a phenomenon that involves changes in the nuclear gene expression in response to
OXPHOS regulatory or structural genes
Altered Gene
Expression
Fig. 2 Mitochondrial-nuclear
intergenomic signaling in
response to OXPHOS
defects. Mitochondrial
OXPHOS dysfunction in
cancer cells can activate a
retrograde signaling cascade.
Altered retrograde signaling
can affect the expression of
various nuclear genes and
cellular pathways that allow
tumors cells to adapt to the
environment and/or promote
tumor cell growth or
progression
Calcium signaling pathway
Redox signaling / antioxidant defense pathway
Cellular morphology and architecture
Tumor invasion or metastatic genes
Nucleus
Retrograde Signaling
OXPHOS Dysfunction
60
S. Srivastava, C.T. Moraes
OXPHOS dysfunction. These nuclear gene expression changes allow the tumor cells
to adapt to the environment that may promote tumor cell growth and/or progression. Studies have shown that a variety of nuclear genes are altered in response
to OXPHOS defects in different tumor cell types and have implicated their role in
tumorigenesis.
3.1 Retrograde Signaling in Cancer Cells
Retrograde signaling has been implicated as an important mechanism in carcinogenesis although the precise mechanism(s) of this pathway is yet elusive in mammalian
systems. Studies have shown that retrograde signaling in response to mitochondrial dysfunction can alter the expression of nuclear genes involved in energy
metabolism, tumor invasion and metastasis, calcium signaling, redox signaling, antiapoptotic genes, as well as genes involved in regulating cellular morphology and
architecture.
3.1.1 Altered Expression of OXPHOS Regulatory Genes and Subunits
The mitochondrial OXPHOS complexes (I, III, IV, and V) comprise subunits
encoded by both mtDNA and nuclear DNA. The expression of nuclear encoded
OXPHOS subunits is under a tight regulation by regulatory genes, namely transcription factors and transcriptional coactivators. Nuclear respiratory factors 1 and
2 (NRF-1 and NRF-2) are the nuclear transcription factors, and the members of
the peroxisome-proliferator activated ␥ coactivator 1 (PGC-1) gene family are
the nuclear transcriptional coactivators that directly or indirectly modulate the
expression of nuclear encoded OXPHOS genes, respectively. Besides regulating the
expression of nuclear encoded OXPHOS genes, the NRFs also exert a regulatory
control over the expression of mitochondrial transcription and replication machinery components, the protein import machinery components, and heme biosynthesis. A change in the expression levels or activity of NRFs and/or PGC-1 family of
transcriptional coactivators would therefore influence the expression of OXPHOS
subunits and mitochondrial biogenesis.
Altered expression of OXPHOS subunits and transcriptional regulatory genes has
been reported in different tumor cell types [22, 23]. We studied a human colorectal
tumor cell line (termed V425) harboring a homoplasmic nonsense mutation in the
mitochondrial COXI gene. Despite the mtDNA mutation load, these cells were found
to maintain a very high rate of mitochondrial respiration (∼75% of control cells) and a
low level of COX activity (∼10% of control cells). Interestingly, osteosarcoma cybrids
harboring the V425 mutant mtDNA showed a significant decline in respiration and
COX activity [22]. Previous studies have shown that ∼10% COX activity cannot sustain a functional respiratory chain [24]. We found that the steady-state transcript levels
of a transcriptional coactivator PGC-1␣ and its homolog PGC-1␤ were highly upregulated in V425 cells. In parallel, we also observed an increase in the steady-state levels
of several mitochondrial proteins (e.g., SDH, COXIV, and cytochrome c) in these
Cellular Adaptations to OXPHOS Defects in Cancer
61
cells. Further, the overexpression of PGC-1␣ in osteosarcoma cybrids harboring the
V425 mutant mtDNA showed a significant improvement in mitochondrial respiration
and an increase in the steady-state levels of several mitochondrial proteins suggesting that PGC-1␣ upregulation can improve mitochondrial respiration in a colorectal
cancer cell line harboring COX deficiency [22]. It has also been shown that PGC-1␣
stimulates mitochondrial biogenesis and respiration by activating the expression of
NRFs, augmenting their transcriptional activity or activating both the expression and
transcriptional activity [25].
Altered expression of PGC-1 related coactivator (PRC), NRF-1, and mitochondrial transcription factor A (TFAM) have also been observed in thyroid oncocytomas
harboring an OXPHOS defect [23]. Savagner et al. found that dense mitochondrial
proliferation in 90% of the tested oncocytic tumors were associated with overexpression of the PRC compared with the controls suggesting that increased mitochondrial
biogenesis might be a feedback mechanism to compensate for the presence of
OXPHOS deficit in these tumors. They also found that the increase in PRC levels was associated with a 5-fold increase in NRF-1 transcripts, 10-fold increase in
TFAM transcripts, and a 3-fold increase in COX activity further suggesting that the
overexpression of PRC pathway is likely responsible for the increased mitochondrial proliferation in thyroid oncocytomas [23].
In contrast with the above studies, low levels of expression of peroxisomeproliferator activated receptor gamma (PPAR-␥) and PGC-1 have also been reported
in breast cancer tissues [26]. Jiang et al. showed that the human metastatic breast
cancer tissues have low levels of PPAR-␥ and PGC-1 transcripts compared with
those of the normal tissues. Further, lower levels of these molecules were found to
be associated with poor clinical outcomes in breast cancer patients [26]. Interestingly, agonists of PPAR-␥ have also been shown to inhibit growth and proliferation
of cancer cells by inducing apoptosis [27].
3.1.2 Activation of Genes Involved in Tumor Invasion and Progression
OXPHOS dysfunction has been shown to modulate tumor invasive properties by
transcriptional upregulation of genes coding for specific members of matrix remodeling pathway and activation of tumor-specific marker genes. van Waveren et al.
showed that OXPHOS dysfunction can alter the expression of nuclear genes coding for factors involved in extracellular matrix remodeling in human osteosarcoma
cells [28]. These genes included members of the matrix metalloproteinases (MMPs)
and tissue inhibitors of metalloproteinases (TIMP) family, urokinase plasminogen
activator and its inhibitor, plasminogen-activator inhibitor 1 (PAI-1), and CTGF
and CYR61 (members of the cysteine-rich 61, connective tissue growth factor and
nephroblastoma-overexpressed [CCN] gene family of growth regulators). Changes
at the protein level for some of these factors have also been observed, though at a
lesser magnitude. The study also showed that osteosarcoma cells harboring mtDNA
mutations were associated with an increased matrigel invasion [28].
Activation of tumor-specific marker genes in response to an OXPHOS dysfunction was observed by Amuthan et al. [12, 13]. They found that partial mtDNA
62
S. Srivastava, C.T. Moraes
depletion (genetic stress) or treatment of cells with mitochondrial-specific inhibitors
(metabolic stress) activated Ca2+ -dependent kinases (PKC, ERK1, ERK2, and calcineurin) and signaling pathways that in turn activated the expression of nuclear
genes involved in tumor invasion and progression. Among these genes, the extracellular matrix protease cathepsin L, transforming growth factor ␤ (TGF-␤), and mouse
melanoma antigen (MMA) were found to be upregulated in both genetically and
metabolically stressed rhabdomyosarcoma and pulmonary carcinoma cells. Using in
vitro Matrigel invasion assays and in vivo rat tracheal xenotransplants in Scid mice,
they further showed that these genetically and metabolically stressed cells were ∼4to 6-fold more invasive than were their respective controls thereby implicating a
role of OXPHOS dysfunction in inducing genes involved in tumor invasion and progression. The invasive behavior of mtDNA-depleted rhabdomyosarcoma and pulmonary carcinoma cells was reverted by the restoration of the mtDNA content [12,
13]. These findings clearly established a role of mitochondrial retrograde signaling in inducing phenotypic changes and tumor progression in osteosarcoma, rhabdomyosarcoma, and human pulmonary carcinoma cells. OXPHOS dysfunction was
also shown to be associated with tumor aggressiveness in renal carcinoma cells [14].
Retrograde signaling in breast cancer has been reported by Delsite et al. They performed a comparative microarray analysis of the nuclear gene expression changes in
a human breast cancer cell line (MDA-435) and its ␳o derivative devoid of mtDNA.
They found that the expression of several nuclear genes involved in cell signaling,
cell growth, energy metabolism, cell architecture, cell differentiation, and apoptosis
were altered in the ␳o derivative of the breast cancer cell line [29]. Functional characterization of these genes is further required to gain a clear insight to the potential
role of retrograde signaling in tumorigenesis.
3.1.3 Activation of Calcium Signaling Pathway
Changes in intracellular calcium (from the approximate 100 nM at rest) mediate a number of processes that affect tumorigenesis, including angiogenesis, motility, cell cycle,
differentiation, transcription, telomerase activity, and apoptosis [30]. Mitochondria
take up Ca2+ in a process dependent on a ⌬⌿m [31]. Ca2+ uptake by mitochondria controls organelle metabolic activity, as pyruvate, ␣-ketoglutarate, and isocitrate dehydrogenases are activated by Ca2+ . In addition, some metabolite transporters have been
shown to be regulated by Ca2+ as well and to enhance aerobic metabolism. Mitochondria, by buffering local [Ca2+ ] (generated by Ca2+ channels on the plasma membrane
or the ER/SR), can also modulate intracellular Ca2+ [31]. Therefore, Ca2+ can regulate cell growth and survival by modulating OXPHOS function. Likewise, a defective OXPHOS impairs Ca2+ buffering. As described above, impaired Ca2+ buffering
capacity by defective mitochondria can activate proteins involved in tumor invasion
and progression, including PKC, ERK1, ERK2, and calcineurin.
OXPHOS dysfunction has been shown to activate calcium-dependent signaling events in several tumor cell types. Avadhani and co-workers investigated the
nuclear gene expression changes in C2C12 rhabdomyosarcoma and human A549
pulmonary carcinoma cells in response to OXPHOS dysfunction and found that
Cellular Adaptations to OXPHOS Defects in Cancer
63
calcium-dependent signaling events were activated in these cells [12, 13, 32].
OXPHOS defects associated with decreased mitochondrial membrane potential
(⌬⌿m) and ATP synthesis caused a sustained increase in intracellular calcium
[Ca2+ ]i levels in these cells. They found that the expression of ryanodine receptor-1
(RyR-1) and ryanodine recptor-2 (RyR-2) genes was upregulated in C2C12 rhabdomyosarcoma and A549 pulmonary carcinoma cells, respectively. Increased RyR
gene expression correlated with high levels of Ca2+ release from ER to the cytosol
in response to caffeine stimulation (RyR agonist). Avadhani and co-workers suggested that the defect in ⌬⌿m was associated with the observed increase in [Ca2+ ]i
levels as mitochondria are key players in sequestering cytosolic Ca2+ and a defect
in ⌬⌿m would affect mitochondrial Ca2+ uptake ability. Restoration of the ⌬⌿m in
both C2C12 rhabdomyosarcoma and A549 pulmonary carcinoma cells was found
to revert the ryanodine receptor gene expression and [Ca2+ ]i levels to near normal
levels suggesting a direct link between OXPHOS dysfunction and elevated [Ca2+ ]i
in these cancer cells. Ca2+ responsive factors such as calcineurin, calcineurindependent NFATc (cytosolic counterpart of activated T-cell–specific nuclear factor), and JNK (c-Jun N-terminal kinase)-dependent ATF2 (activated transcription
factor 2) were also found to be elevated in both C2C12 rhabdomyosarcoma and
pulmonary carcinoma cells. Further, Ca2+ -dependent PKC (protein kinase C) and
MAPK (mitogen activated protein kinase) genes were activated in C2C12 rhabdomyosarcoma and lung carcinoma cells, respectively. Another study by Avadhani
and co-workers has shown that calcineurin that is activated in response to elevated
cytosolic Ca2+ inactivates I-␬B␤ thereby leading to an activation of the NF-␬B/Rel
family of transcription factors in the OXPHOS-deficient C2C12 rhabdomyosarcoma
cells [33]. They suggested that the NF-␬B/Rel family of transcription factors might
be involved in the activation of nuclear genes in response to OXPHOS dysfunction
in these cells [33]. Rise in cytosolic Ca2+ in response to OXPHOS dysfunction has
also been observed in rat pheochromocytoma PC12 cells. Luo et al. showed that
the treatment of PC12 cells with mitochondrial uncoupler FCCP releases Ca2+ from
internal stores leading to a rise in cytosolic Ca2+ , which in turn activates MAPK
also suggesting a potential link between OXPHOS dysfunction and Ca2+ signaling
in these cells [34].
3.1.4 Activation of Redox Signaling Pathway
Changes in nuclear gene expression in response to oxidative stress have been
observed in a variety of tumor cells. mtDNA mutations or OXPHOS dysfunction
predispose mitochondria to the generation of ROS leading to oxidative stress. Cellular oxidative stress has been implicated in initiation, promotion, and progression
of carcinogenesis [35]. Low levels of ROS generation have been shown to be mitogenic in a variety of human and mouse cell types [36, 37]. For example, under cell
culture conditions, 10 nM to 1 ␮M concentrations of both O2 •- and H2 O2 have been
shown to stimulate the growth of hamster and rat fibroblasts [37]. Low levels of
ROS were also shown to stimulate the growth of mouse epidermal cells, human
fibroblasts, and human astrocytoma cells [38, 39]. The increased tumorigenesis of
64
S. Srivastava, C.T. Moraes
NARP cells with a pathogenic mtDNA mutation described above [10] could also be
related to increased ROS, as the mutation in ATPase 6 was reported to significantly
increase ROS production [40]. Further, many chemical carcinogens initiate carcinogenesis by stimulating ROS production, which in turn activates the expression of
early growth related genes such as c-fos, c-myc, and c-jun [41].
Although the molecular mechanism(s) of ROS-mediated cell growth are yet
unclear, cellular ROS have been shown to activate the expression of redox-sensitive
nuclear genes and signaling pathways. It has been shown that ROS activate the
expression of nuclear factor kappa B (NF-␬B) and activator protein (AP-1) transcription factors [42, 43]. Both NF-␬B and AP-1 are redox-sensitive transcription
factors. NF-␬B in its active form consists of two subunits, p50 and p65. In the
absence of a stimulus, it is present in the cytoplasm in association with its inhibitory
subunit I-␬B. In response to a stimulus that leads to the I-␬B subunit phosphorylation, NF-␬B dissociates from the I-␬B inhibitory complex and translocates to the
nucleus where it activates the expression of specific target genes [44]. AP-1 is a heterodimer of transcription factors composed of Jun, Fos, or activating transcription
factor (ATF) subunits and binds the AP-1 sites in the promoter region to activate the
target gene expression [45].
Constitutive activation of NF-␬B and Ap-1 genes in response to oxidative stress
has been observed in a variety of cancer cell types. Shi et al. demonstrated that
OH radicals activated the expression of NF-␬B in Jurkat cells, macrophages, and
mouse epidermal cells, and antioxidants that scavenge OH radicals inhibited its activation [46]. High levels of ROS were also shown to cause a constitutive elevation of
NF-␬B and AP-1 transcription factors in transformed mouse keratinocyte cells [7].
Constitutive elevation of AP-1 activity has also been shown to be associated with
the conversion of benign papillomas to malignant carcinomas in mouse epidermal
cells suggesting that the target genes induced by activated AP-1 were involved in
cell growth and metastasis [47]. Inhibition of Ap-1 activity by a dominant negative
c-Jun mutant was shown to suppress the tumorigenic phenotype of the malignant
mouse epidermal cells thereby blocking the tumor formation in nude mice [48].
Constitutive activation of NF-␬B has also been shown to induce genes involved
in tumor progression and metastasis [49]. Further, treatment with an antioxidant,
N-acetylcysteine (NAC), has been shown to inhibit the elevated expression of NF␬B and AP-1 levels in tumor cells implicating the role of ROS in the activation of
redox-sensitive signaling pathways [7].
The activities of NF-␬B and AP-1 transcription factors are modulated by phosphorylation. Although the precise signaling events that activate the phosphorylation of these transcription factors in response to increase in cellular ROS levels are
yet not fully understood, protein kinases are central in their activation. It has been
observed that MAPK pathways are involved in signaling the activation of both NF␬B and Ap-1 transcription factors. There are three subtypes of MAPKs; the extracellular signal regulated kinases (ERKs), the c-jun N-terminal kinases (JNKs), and
the p38 MAPKS. Gupta et al. showed that the NF-␬B activation in malignantly progressed mouse keratinocytes was associated with an increase in ERK-1/2 and p38
MAPK activities and that NAC treatment rapidly abolished the increase in MAPK
activities [7]. Direct activation of I-␬B by MEKK1 (MAPK/ERK kinase kinase-1)
Cellular Adaptations to OXPHOS Defects in Cancer
65
has also been demonstrated in vitro [50]. Activation of AP-1 by MAPK in response
to various stimuli has been demonstrated [51]. There is also evidence that antioxidants can attenuate MAPK activation suggesting that MAPK signaling cascades are
activated in response to cellular oxidative stress conditions [7].
3.1.5 Altered Antioxidant Defense Pathway
The antioxidant defense enzymes have been found to be altered in several human
tumors. Superoxide dismutases (SODs) are the first line of cellular defense against
increase in intracellular ROS. There are two forms of intracellular SODs: a
manganese-containing SOD (MnSOD) that localizes to the mitochondria and a
copper-zinc–containing SOD that localizes to the cytosol. Most human tumor cell
types have been found to be deficient in their antioxidant defense function. Antioxidant enzymes including catalase and glutathione peroxidase have also been shown
to be lowered in cancer cells [52, 53] suggesting that tumor cells have a deficient
antioxidant function.
Studies have shown that antioxidant treatment prevents the malignant transformation of cancer cells. Increased expression of MnSOD has been shown to suppress cancer phenotypes in several murine and human cancer cells. St Clair et al.
showed that overexpression of human MnSOD in mouse cells significantly reduced
the frequency of radiation-induced neoplastic transformation implicating a direct
link between mitochondrial antioxidants and neoplastic transformation [54]. Zhao
et al. demonstrated that in response to tumor promoters (phorbol esters), transgenic
mice expressing human MnSOD in the skin showed a significant reduction in papilloma formation compared with that of the non-transgenic controls [55]. They further found that the decreased papilloma formation was associated with a delay and
reduction in AP-1 transcription factor binding activity suggesting that the MnSOD
overexpression suppressed the tumor formation and Ap-1 activation in these transgenic animals [55]. Overexpression of human MnSOD has also been shown to
suppress the malignant phenotype of human pancreatic and breast cancer cells. Li
et al. demonstrated that MnSOD-overexpressing MCF-7 breast cancer cells showed
a marked inhibition in growth rate in vitro under culture conditions and in vivo upon
inoculation in nude mice when compared with the wild-type MCF-7 cells [56]. In
another study, they demonstrated that the transcriptional and DNA binding ability of
NF-␬B and AP-1 were reduced by ∼50% in MnSOD-overexpressing MCF-7 cells
[57]. They further showed that the expression of NF-␬B and AP-1 responsive genes
was downregulated in these cells when compared with that of the wild-type MCF-7
cells suggesting that the tumor-suppressive effects of MnSOD were associated with
the inhibition of NF-␬B and Ap-1 activities that in turn downregulated the genes
involved in malignant progression [57].
3.1.6 Altered Mitochondrial Morphology, Cell Surface, and Architecture
Mitochondrial fission and fusion are the central events that regulate mitochondrial
morphology. Changes in mitochondrial morphology appear to be a key event during
retrograde signaling. Studies have implicated a role of mitochondrial morphology in
66
S. Srivastava, C.T. Moraes
modulating respiratory activity. For example, downregulation of mitofusin 2 (Mfn 2)
decreases mitochondrial respiration and increases the cellular glucose uptake levels
for glycolysis. Further, ectopic upregulation of Mfn 2 increases the expression of
OXPHOS subunits, glucose oxidation, and ⌬⌿m [58]. Optic Atrophy 1 (Opa 1) is
an important player in the formation of mitochondrial cristae. Downregulation of
Opa 1 has been shown to induce the loss of mitochondrial respiration and increased
mitochondrial fragmentation [59].
Changes in mitochondrial morphology and ultrastructure were reported in different tumor cell types over several decades. Tumor cell mitochondria were found
to contain abnormal size, shape (dumbbell or cup), cristae organization, or inclusions. For example, rapidly growing hepatomas were found to contain small-size
mitochondria with fewer cristae, whereas slowly growing hepatomas were found to
contain large-size mitochondria with densely packed cristae [60]. Changes in the
mitochondrial morphology in tumor cells could therefore affect their respiratory
activity. Conversely, it is also possible that in response to altered respiratory activity
or OXPHOS dysfunction, the retrograde signaling events in tumor cells affect the
expression of nuclear genes involved in the regulation of mitochondrial morphology
and ultrastructure.
Changes in cell morphology and architecture are a characteristic feature of many
cancer cells. These changes occur in cell surface charge, glycoprotein composition,
membrane organization, and/or the ability of cells to agglutinate on lectin. The initial evidence of a potential role of OXPHOS dysfunction in inducing cell surface
changes was obtained in yeast. In 1980, Evans and colleagues showed that mitochondrial petite mutations in yeast that cause respiratory deficiency lead to drastic
changes in cell surface charge, cell wall organization, and lectin agglutinability in
contrast with the wild-type yeast cells suggesting that the OXPHOS deficiency in
yeast altered the retrograde signaling pathway leading to gene expression changes
and hence altered cell surface properties [61]. Preliminary evidence in literature has
also suggested a link between OXPHOS dysfunction and cell surface properties in
mammalian cells. For example, Soslau et al. showed that baby hamster kidney cells
when treated with ethidium bromide (Etbr) show an altered glycoprotein composition in the plasma membrane [62]. Further, they also showed that the glycopeptide
elution profile of these Etbr-treated cells was similar to that of the cells transformed
with Rous sarcoma virus. Low concentrations of Etbr are known to intercalate to
mtDNA and abrogate mtDNA replication. Exponentially growing cells when treated
with Etbr (25 ng/mL to 2 ␮g/mL) undergo mtDNA loss leading to either partial or
complete loss of mtDNA molecules, generating an OXPHOS defect. These findings therefore suggest that altered retrograde signaling events during OXPHOS dysfunction could activate nuclear genes or pathways that cause global changes in cell
surface properties.
3.1.7 Activation of Antiapoptotic Genes
Activation of antiapoptotic genes is a common occurrence in most human cancers. However, retrograde signaling pathway has also been shown to activate
Cellular Adaptations to OXPHOS Defects in Cancer
67
antiapoptotic genes (e.g., Bcl-xL and Bcl-2) in certain cancer cell types [12, 63].
Increased expression of the antiapoptotic genes may also contribute to metabolic
homeostasis. It has been shown that members of the Bcl-2 antiapoptotic family of
proteins regulate ATP/ADP exchange across the mitochondrial membranes and can
prevent the loss of mitochondrial respiration during apoptosis [64]. Manfredi and
colleagues showed that Bcl-2/Bcl-xL can improve OXPHOS function in cells harboring pathogenic mtDNA mutations. The effect of Bcl-2 overexpression in mutated
cells was found to be independent from apoptosis and was suggested to modulate the
adenine nucleotide exchange between mitochondria and cytosol. This study therefore provided evidence that when OXPHOS function is reduced, Bcl-2 expression
can improve it by regulating the levels of adenine nucleotides inside the mitochondria [64].
3.2 Mechanism of Retrograde Signaling
The mechanism of retrograde signaling was first identified in yeast and it is still
poorly understood in other systems. Three retrograde regulatory genes (RTG1,
RTG2, and RTG3) play a central role in retrograde signaling in yeast [65] (Fig. 3).
Rtg1p and Rtg3p are transcription factors, whereas Rtg2p is mitochondrial function
sensor. During normal mitochondrial function, Rtg1p and Rtg3p form heterodimers
P
Nucleus
Rtg3p
Rtg1p
Target gene
GTCAC
R box
P
Rtg3p
P P
P Rtg3p
P
Rtg1p
Rtg2p
Rtg1p
Sensor
Mitochondrial
Dysfunction
Cytoplasm
Fig. 3 Schematic model of retrograde signaling in yeast. RTG genes play a central role in retrograde signaling. Rtg1p and Rtg3p are sequestered in cytoplasm during normal mitochondrial
function. Rtg3p is highly phosphorylated during this state. In response to mitochondrial dysfunction (sensed by Rtg2p), the Rtg3p undergoes partial dephosphorylation. Both Rtg3p and Rtg1p
translocate to the nucleus and activate target gene expression leading to changes in nuclear genes
and/or signaling pathways that in turn can alter cellular function and homeostasis
68
S. Srivastava, C.T. Moraes
and are sequestered in the cytoplasm. Rtg3p is phosphorylated at multiple sites
during this state. In response to mitochondrial dysfunction (sensed by Rtg2p), a
retrograde signaling pathway gets activated. The Rtg3p undergoes partial dephosphorylation and translocates to the nucleus. Rtg1p also follows and translocates to
the nucleus. Both Rtg1p and Rtg3p activate transcription at the target genes. In the
absence of Rtg2p, the Rtg1p/Rtg3p complex remains in the cytoplasm during active
retrograde signaling pathway implicating its role as mitochondrial function sensor
and in signal relay to the Rtg1p/Rtg3p complex [65].
The precise mechanism of retrograde signaling in mammalian systems is yet
not known. The mammalian homologs of RTG genes have yet not been identified.
However, human MYC protein has been found to share a significant homology
with yeast Rtg3p, and it is involved in the regulation of many cellular functions
such as regulation of glycolysis, cellular stress response cell cycle progression, and
apoptosis [66, 67].
4 Conclusion
Mitochondrial dysfunction has been reported in a variety of human cancers.
Mitochondrial-nuclear intergenomic signaling is an important phenomenon in regulating mitochondrial and cellular function and homeostasis. Although the role of
mtDNA mutations or OXPHOS defects in cancer progression is poorly understood,
a considerable body of evidence has now shown that mitochondrial function has an
important role in different aspects of tumorigenesis. Processes such as cell survival
and cell invasion can be directly influenced by OXPHOS function. Less clear is
the association between cell cycle progression and an OXPHOS defect. However,
specific examples, such as SDH mutations in paragangliomas and pheochromocytomas, provide proof of principle that OXPHOS defects can somehow stimulate cell
division.
It appears likely that some properties related to OXPHOS dysfunction can have
an advantageous consequence for tumor development. However, this feature could
impair cell performance at a different level. Therefore, tumor cells with an OXPHOS
defect, in some cases, have compensatory mechanisms, such as an increase in mitochondrial biogenesis via overexpression of PGC-1 family members. Studies on the
interplay between these metabolic and signaling pathways will help us better understand the basic mechanisms of tumor progression.
Note
A recent study by Ishikawa et al. (Ishikawa K, Takenaga K, Akimoto M, et al.
ROS-generating mitochondrial DNA mutations can regulate tumor cell metastasis.
Cellular Adaptations to OXPHOS Defects in Cancer
69
Science. 2008;320(5876):661–664) showed that specific mtDNA changes in complex I subunit can increase the metastatic properties of tumor cells.
References
1. Warburg O. On the origin of cancer cells. Science 1956; 123(3191):309–314.
2. Pedersen PL, Greenawalt JW, Chan TL, Morris HP. A comparison of some ultrastructural and
biochemical properties of mitochondria from Morris hepatomas 9618A, 7800, and 3924A.
Cancer Res 1970; 30(11):2620–2626.
3. Polyak K, Li Y, Zhu H, et al. Somatic mutations of the mitochondrial genome in human
colorectal tumours. Nat Genet 1998; 20(3):291–293.
4. Fliss MS, Usadel H, Caballero OL, et al. Facile detection of mitochondrial DNA mutations in
tumors and bodily fluids. Science 2000; 287(5460):2017–2019.
5. Liu VW, Shi HH, Cheung AN, et al. High incidence of somatic mitochondrial DNA mutations
in human ovarian carcinomas. Cancer Res 2001; 61(16):5998–6001.
6. Nagy A, Wilhelm M, Sukosd F, Ljungberg B, Kovacs G. Somatic mitochondrial DNA mutations in human chromophobe renal cell carcinomas. Genes Chromosomes Cancer 2002;
35(3):256–260.
7. Gupta A, Rosenberger SF, Bowden GT. Increased ROS levels contribute to elevated transcription factor and MAP kinase activities in malignantly progressed mouse keratinocyte cell lines.
Carcinogenesis 1999; 20(11):2063–2073.
8. Irani K, Xia Y, Zweier JL, et al. Mitogenic signaling mediated by oxidants in Ras-transformed
fibroblasts. Science 1997; 275(5306):1649–1652.
9. Coller HA, Khrapko K, Bodyak ND, Nekhaeva E, Herrero-Jimenez P, Thilly WG. High frequency of homoplasmic mitochondrial DNA mutations in human tumors can be explained
without selection. Nat Genet 2001; 28(2):147–150.
10. Petros JA, Baumann AK, Ruiz-Pesini E, et al. mtDNA mutations increase tumorigenicity in
prostate cancer. Proc Natl Acad Sci USA 2005; 102(3):719–724.
11. Shidara Y, Yamagata K, Kanamori T, et al. Positive contribution of pathogenic mutations in
the mitochondrial genome to the promotion of cancer by prevention from apoptosis. Cancer
Res 2005; 65(5):1655–1663.
12. Amuthan G, Biswas G, Ananadatheerthavarada HK, Vijayasarathy C, Shephard HM, Avadhani NG. Mitochondrial stress-induced calcium signaling, phenotypic changes and invasive
behavior in human lung carcinoma A549 cells. Oncogene 2002; 21(51):7839–7849.
13. Amuthan G, Biswas G, Zhang SY, Klein-Szanto A, Vijayasarathy C, Avadhani NG.
Mitochondria-to-nucleus stress signaling induces phenotypic changes, tumor progression and
cell invasion. EMBO J 2001; 20(8):1910–1920.
14. Simonnet H, Alazard N, Pfeiffer K, et al. Low mitochondrial respiratory chain content
correlates with tumor aggressiveness in renal cell carcinoma. Carcinogenesis 2002; 23(5):
759–768.
15. Simonnet H, Demont J, Pfeiffer K, et al. Mitochondrial complex I is deficient in renal oncocytomas. Carcinogenesis 2003; 24(9):1461–1466.
16. Savagner F, Franc B, Guyetant S, Rodien P, Reynier P, Malthiery Y. Defective mitochondrial
ATP synthesis in oxyphilic thyroid tumors. J Clin Endocrinol Metab 2001; 86(10):4920–4925.
17. Akimoto M, Niikura M, Ichikawa M, et al. Nuclear DNA but not mtDNA controls tumor
phenotypes in mouse cells. Biochem Biophys Res Commun 2005; 327(4):1028–1035.
18. Tomlinson IP, Alam NA, Rowan AJ, et al. Germline mutations in FH predispose to dominantly
inherited uterine fibroids, skin leiomyomata and papillary renal cell cancer. Nat Genet 2002;
30(4):406–410.
19. Astuti D, Latif F, Dallol A, et al. Gene mutations in the succinate dehydrogenase subunit
SDHB cause susceptibility to familial pheochromocytoma and to familial paraganglioma. Am
J Hum Genet 2001; 69(1):49–54.
70
S. Srivastava, C.T. Moraes
20. Niemann S, Muller U. Mutations in SDHC cause autosomal dominant paraganglioma, type 3.
Nat Genet 2000; 26(3):268–270.
21. Baysal BE, Ferrell RE, Willett-Brozick JE, et al. Mutations in SDHD, a mitochondrial complex II gene, in hereditary paraganglioma. Science 2000; 287(5454):848–851.
22. Srivastava S, Barrett JN, Moraes CT. PGC-1alpha/beta upregulation is associated with
improved oxidative phosphorylation in cells harboring nonsense mtDNA mutations. Hum Mol
Genet 2007; 16(8):993–1005.
23. Savagner F, Mirebeau D, Jacques C, et al. PGC-1-related coactivator and targets are upregulated in thyroid oncocytoma. Biochem Biophys Res Commun 2003; 310(3):779–784.
24. Villani G, Greco M, Papa S, Attardi G. Low reserve of cytochrome c oxidase capacity in
vivo in the respiratory chain of a variety of human cell types. J Biol Chem 1998; 273(48):
31829–31836.
25. Wu Z, Puigserver P, Andersson U, et al. Mechanisms controlling mitochondrial biogenesis
and respiration through the thermogenic coactivator PGC-1. Cell 1999; 98(1):115–124.
26. Jiang WG, Douglas-Jones A, Mansel RE. Expression of peroxisome-proliferator activated
receptor-gamma (PPARgamma) and the PPARgamma co-activator, PGC-1, in human breast
cancer correlates with clinical outcomes. Int J Cancer 2003; 106(5):752–757.
27. Ohta K, Endo T, Haraguchi K, Hershman JM, Onaya T. Ligands for peroxisome proliferatoractivated receptor gamma inhibit growth and induce apoptosis of human papillary thyroid
carcinoma cells. J Clin Endocrinol Metab 2001; 86(5):2170–2177.
28. van Waveren C, Sun Y, Cheung HS, Moraes CT. Oxidative phosphorylation dysfunction
modulates expression of extracellular matrix–remodeling genes and invasion. Carcinogenesis 2006; 27(3):409–418.
29. Delsite R, Kachhap S, Anbazhagan R, Gabrielson E, Singh KK. Nuclear genes involved in
mitochondria-to-nucleus communication in breast cancer cells. Mol Cancer 2002; 1(1):6.
30. Monteith GR, McAndrew D, Faddy HM, Roberts-Thomson SJ. Calcium and cancer: targeting
Ca2+ transport. Nat Rev Cancer 2007; 7(7):519–530.
31. Giacomello M, Drago I, Pizzo P, Pozzan T. Mitochondrial Ca2+ as a key regulator of cell life
and death. Cell Death Differ 2007; 14(7):1267–1274.
32. Biswas G, Adebanjo OA, Freedman BD, et al. Retrograde Ca2+ signaling in C2C12 skeletal
myocytes in response to mitochondrial genetic and metabolic stress: a novel mode of interorganelle crosstalk. EMBO J 1999; 18(3):522–533.
33. Biswas G, Anandatheerthavarada HK, Zaidi M, Avadhani NG. Mitochondria to nucleus stress
signaling: a distinctive mechanism of NFkappaB/Rel activation through calcineurin-mediated
inactivation of IkappaBbeta. J Cell Biol 2003; 161(3):507–519.
34. Luo Y, Bond JD, Ingram VM. Compromised mitochondrial function leads to increased cytosolic calcium and to activation of MAP kinases. Proc Natl Acad Sci USA 1997; 94(18):
9705–9710.
35. Cerutti PA. Prooxidant states and tumor promotion. Science 1985; 227(4685):375–381.
36. Burdon RH, Rice-Evans C. Free radicals and the regulation of mammalian cell proliferation.
Free Radic Res Commun 1989; 6(6):345–358.
37. Burdon RH, Gill V, Rice-Evans C. Oxidative stress and tumour cell proliferation. Free Radic
Res Commun 1990; 11(1–3):65–76.
38. Arora-Kuruganti P, Lucchesi PA, Wurster RD. Proliferation of cultured human astrocytoma
cells in response to an oxidant and antioxidant. J Neurooncol 1999; 44(3):213–221.
39. Murrell GA, Francis MJ, Bromley L. Modulation of fibroblast proliferation by oxygen free
radicals. Biochem J 1990; 265(3):659–665.
40. Mattiazzi M, Vijayvergiya C, Gajewski CD, et al. The mtDNA T8993G (NARP) mutation
results in an impairment of oxidative phosphorylation that can be improved by antioxidants.
Hum Mol Genet 2004; 13(8):869–879.
41. Amstad P, Crawford D, Muehlematter D, Zbinden I, Larsson R, Cerutti P. Oxidants stress
induces the proto-oncogenes, C-fos and C-myc in mouse epidermal cells. Bull Cancer 1990;
77(5):501–502.
Cellular Adaptations to OXPHOS Defects in Cancer
71
42. Abate C, Patel L, Rauscher FJ 3rd, Curran T. Redox regulation of fos and jun DNA-binding
activity in vitro. Science 1990; 249(4973):1157–1161.
43. Flohe L, Brigelius-Flohe R, Saliou C, Traber MG, Packer L. Redox regulation of NF-kappa
B activation. Free Radic Biol Med 1997; 22(6):1115–1126.
44. Baeuerle PA. The inducible transcription activator NF-kappa B: regulation by distinct protein
subunits. Biochim Biophys Acta 1991; 1072(1):63–80.
45. Karin M, Liu Z, Zandi E. AP-1 function and regulation. Curr Opin Cell Biol 1997; 9(2):
240–246.
46. Shi X, Dong Z, Huang C, et al. The role of hydroxyl radical as a messenger in the activation
of nuclear transcription factor NF-kappaB. Mol Cell Biochem 1999; 194(1–2):63–70.
47. Domann FE Jr, Levy JP, Finch JS, Bowden GT. Constitutive AP-1 DNA binding and transactivating ability of malignant but not benign mouse epidermal cells. Mol Carcinog 1994;
9(2):61–66.
48. Domann FE, Levy JP, Birrer MJ, Bowden GT. Stable expression of a c-JUN deletion mutant in
two malignant mouse epidermal cell lines blocks tumor formation in nude mice. Cell Growth
Differ 1994a; 5(1):9–16.
49. Nakshatri H, Bhat-Nakshatri P, Martin DA, Goulet RJ Jr, Sledge GW Jr. Constitutive activation of NF-kappaB during progression of breast cancer to hormone-independent growth. Mol
Cell Biol 1997; 17(7):3629–3639.
50. Lee FS, Hagler J, Chen ZJ, Maniatis T. Activation of the IkappaB alpha kinase complex by
MEKK1, a kinase of the JNK pathway. Cell 1997; 88(2):213–222.
51. Karin M. The regulation of AP-1 activity by mitogen-activated protein kinases. J Biol Chem
1995; 270(28):16483–16486.
52. Gupta A, Butts B, Kwei KA, et al. Attenuation of catalase activity in the malignant phenotype plays a functional role in an in vitro model for tumor progression. Cancer Lett 2001;
173(2):115–125.
53. Oberley LW, Oberley TD. Role of antioxidant enzymes in cell immortalization and transformation. Mol Cell Biochem 1988; 84(2):147–153.
54. St Clair DK, Wan XS, Oberley TD, Muse KE, St. Clair WH. Suppression of radiation-induced
neoplastic transformation by overexpression of mitochondrial superoxide dismutase. Mol Carcinog 1992; 6(4):238–242.
55. Zhao Y, Xue Y, Oberley TD, et al. Overexpression of manganese superoxide dismutase suppresses tumor formation by modulation of activator protein-1 signaling in a multistage skin
carcinogenesis model. Cancer Res 2001; 61(16):6082–6088.
56. Li JJ, Oberley LW, St. Clair DK, Ridnour LA, Oberley TD. Phenotypic changes induced in
human breast cancer cells by overexpression of manganese-containing superoxide dismutase.
Oncogene 1995; 10(10):1989–2000.
57. Li JJ, Oberley LW, Fan M, Colburn NH. Inhibition of AP-1 and NF-kappaB by manganesecontaining superoxide dismutase in human breast cancer cells. FASEB J 1998; 12(15):
1713–1723.
58. Pich S, Bach D, Briones P, et al. The Charcot-Marie-Tooth type 2A gene product, Mfn2,
up-regulates fuel oxidation through expression of OXPHOS system. Hum Mol Genet 2005;
14(11):1405–1415.
59. Chen H, Chomyn A, Chan DC. Disruption of fusion results in mitochondrial heterogeneity
and dysfunction. J Biol Chem 2005; 280(28):26185–26192.
60. Hruban Z, Swift H, Rechcigl M Jr. Fine structure of transplantable hepatomas of the rat. J Natl
Cancer Inst 1965; 35(3):459–495.
61. Evans IH, Diala ES, Earl A, Wilkie D. Mitochondrial control of cell surface characteristics in
Saccharomyces cerevisiae. Biochim Biophys Acta 1980; 602(1):201–206.
62. Soslau G, Fuhrer JP, Nass MM, Warren L. The effect of ethidium bromide on the membrane
glycopeptides in control and virus-transformed cells. J Biol Chem 1974; 249(10):3014–3020.
63. Dey R, Moraes CT. Lack of oxidative phosphorylation and low mitochondrial membrane
potential decrease susceptibility to apoptosis and do not modulate the protective effect of
Bcl-x(L) in osteosarcoma cells. J Biol Chem 2000; 275(10):7087–7094.
72
S. Srivastava, C.T. Moraes
64. Manfredi G, Kwong JQ, Oca-Cossio JA, et al. BCL-2 improves oxidative phosphorylation
and modulates adenine nucleotide translocation in mitochondria of cells harboring mutant
mtDNA. J Biol Chem 2003; 278(8):5639–5645.
65. Sekito T, Thornton J, Butow RA. Mitochondria-to-nuclear signaling is regulated by the
subcellular localization of the transcription factors Rtg1p and Rtg3p. Mol Biol Cell 2000;
11(6):2103–2115.
66. Miceli MV, Jazwinski SM. Common and cell type-specific responses of human cells to mitochondrial dysfunction. Exp Cell Res 2005; 302(2):270–280.
67. Epstein CB, Waddle JA, Hale WT, et al. Genome-wide responses to mitochondrial dysfunction. Mol Biol Cell 2001; 12(2):297–308.
Regulation of Glucose and Energy Metabolism
in Cancer Cells by Hypoxia Inducible Factor 1
´
´ C´esar Ferreira and Elida
Tulio
Geralda Campos
Abstract A crucial aspect of carcinogenesis is the adaptation of cells to the changing environmental conditions of the tumor mass. To grow and proliferate, the cells
need nutrients that have to be delivered by surrounding vessels. In addition, they
have to cope with a gradient of oxygen tension as the tumor mass grows. Hypoxia
inducible factor 1 (HIF-1) mediates important responses to physiologic and pathologic processes that allow angiogenesis, erythropoiesis, cell proliferation and survival, and metabolic changes to take place. It is a heterodimeric transcription factor
regulated by a complex network. Among the major metabolic changes evident in
cancer cells is the activation of the glycolytic pathway, even in the presence of oxygen. HIF-1 regulates glucose and energy metabolism by directly activating the gene
expression of glycolytic enzymes, regulatory enzymes, and glucose transporters.
HIF-1␣ (the regulated subunit of the HIF-1 complex) is regulated mainly at the
posttranscriptional level through mechanisms that block its proteasomal degradation
depending on the oxygen concentration. HIF-1 is overexpressed in many tumors,
and evidence is mounting that HIF-1 regulation of glucose and energy metabolism
is important in carcinogenesis and tumor progression.
Keywords HIF-1 · HIF-1␣ · Glycolysis · von Hippel–Lindau tumor suppressor
protein (pVHL) · Lactate · Glucose transporter · Regulation of gene expression ·
Hypoxia response element (HRE) · Prolyl hydroxylases
1 Oxygen Homeostasis
Cells respond to extracellular and intracellular stimuli to maintain homeostasis, and
hypoxia is one of the most fundamental environmental stimuli. Oxygen homeostasis is crucial to maintain the approximately 1014 cells in the adult human body with
´
E.G.
Campos (B)
Departamento de Biologia Celular, Universidade de Bras´ılia, CEP 70910-900, Bras´ılia, DF, Brazil
e-mail: elida@unb.br
S.P. Apte, R. Sarangarajan (eds.), Cellular Respiration and Carcinogenesis,
DOI 10.1007/978-1-59745-435-3 6,
C Humana Press, a part of Springer Science+Business Media, LLC 2009
73
74
´
T.C. Ferreira, E.G.
Campos
an adequate supply of O2 . The maintenance of the O2 levels within strict limits is
essential throughout life but is particularly important during periods of rapid cellular
proliferation, such as in normal growth or during the development of neoplasias. In
normal tissues, the average oxygen tension is 40 to 50 mm Hg, whereas in tumors
these levels are 5 to 10 mm Hg. Control of O2 levels in the tissues of animals
occurs through the combination of gene regulation and biochemical and physiologic mechanisms. Low levels of oxygen in the environment lead cells and tissues
to adapt to this adverse condition. Significant changes in gene expression and activation of signaling pathways occurs in response to hypoxia, a characteristic condition
in solid tumors, resulting in an elevated transcription of angiogenic, hematopoietic,
and some metabolic genes. Hypoxia inducible factor 1 (HIF-1) was discovered in
1992 by Gregg L. Semenza and Guang L. Wang at the Johns Hopkins University
School of Medicine. It is currently known as the central regulator of the response
to low levels of oxygen. HIF-1 is a heterodimeric transcription factor and regulates
genes whose products participate in glycolysis, glucose transport, angiogenesis, cell
proliferation and survival, pH regulation, and cell migration.
2 HIF-1 Structure
HIF-1␣ protein is one of the subunits that form HIF-1 and belongs to the basic
helix-loop-helix (bHLH)-PAS family of transcription factors. PAS is an acronym
that refers to the Drosophila proteins Period (or PER), ARNT, and Single-minded
or (SIM), the first proteins in which this motif was identified. HIF-1␣ has the bHLHPAS domain, which is in the N-terminal region of the protein and has two repeats
(A and B) in its structure. This subunit binds to the second subunit that forms
HIF-1, the aryl hydrocarbon receptor nuclear translocator (ARNT), also referred
as HIF-1␤. ARNT is a nuclear protein constitutively expressed that also participates in the transcriptional response to xenobiotics when interacting with the aryl
hydrocarbon receptor (AHR). As with HIF-1␣, the ␤ subunit also belongs to the
bHLH-PAS family of transcription factors. HIF-1␣ and HIF-1␤ share some amino
acid sequence similarities. The bHLH-PAS domains in both subunits are crucial for
the heterodimerization and binding of these proteins to DNA.
HIF-1␣ is an 826-amino-acid protein that has two transactivation domains
(TADs), one in the N-terminal region (TAD-N) that overlaps with the oxygen degradation dependent (ODD) domain and another in the C-terminal region (TAD-C), the
latter also found in the HIF-1␤ (Fig. 1). The TAD domains are localized at amino
acids 531 to 575 (TAD-N) and 786 to 826 (TAD-C) of the HIF-1␣. The TAD-C
domain of HIF-1␣ interacts with the transcriptional activators CREB binding protein (CBP) and p300. Maintenance of cysteine 800 in the TAD-C in a reduced state
is essential for p300/CBP binding. Therefore, HIF-1␣ redox state is involved in its
activation. HIF-1␣ and p300 interaction requires a leucine-rich hydrophobic interface that is regulated by the reversible change between hydrophobic and hydrophilic
Cys800 of HIF-1␣. The reduced state of this cysteine is maintained by redox factor-1
Regulation of Glucose and Energy Metabolism in Cancer Cells by HIF-1
75
Fig. 1 HIF-1 proteins ␣ and ␤ have similar structural domains. HIF-1␣ (upper drawing) is a
826-amino-acid protein with a basic helix-loop-helix-PAS (bHLH-PAS) domain in the N-terminal
and repeats A and B. There are two transactivation domains (TADs), one located in the N-terminal
region (TAD-C), which overlaps with the oxygen degradation dependent (ODD) domain, and other
in the C-terminal region (TAD-C). Both TADs are connected by an inhibitory domain (ID). HIF1␤ (lower drawing) is a 789-amino-acid protein, which also belongs to the bHLH-PAS family of
transcription factors. As with HIF-1␣, HIF-1␤ also has the bHLH-PAS domain in the N-terminal
and repeats A and B. However, it has only one TAD, which is located in the C-terminal of the
protein (see Color Insert)
[1] and mediated by thioredoxin (TRX). Therefore, compounds that inhibit thioredoxin cause inhibition of HIF-1␣–mediated transactivation of HIF-1 target genes.
Both TAD-C and TAD-N domains are connected by a inhibitory domain (ID). The
ODD domain is a proline-serine-threonine–rich domain involved in the regulation
of protein stability as a function of O2 concentration. In HIF-1␣, the main target
amino acids for hydroxylation are localized in this domain (described below). HIF1␣ protein contains two nuclear localization signals (NLS), one in the N-terminal
(NLS-N) and the other in the C-terminal, which encompass the amino acids 17 to 23
and 718 to 721, respectively [1].
3 HIF-1␣ Regulation
HIF-1␣ regulation is complex and is governed by posttranslational mechanisms
(Fig. 2). Protein stability and activity are regulated by covalent modifications such
as hydroxylation, acetylation, S-nitrosation, and phosphorylation. Among them,
hydroxylation and phosphorylation are the most studied. In normoxia conditions,
hydroxylated HIF-1␣ is recognized by the von Hippel–Lindau tumor suppressor
protein (pVHL), which is part of a polyubiquitination complex. HIF-1␣ is hydroxylated by a HIF prolyl hydroxylase (PHD). pVHL recognizes the hydroxylated proline residues (Pro402 and Pro564) located in the human HIF-1␣ ODD domain.
PHD is a dioxygenase that uses O2 and 2-oxoglutarate (2-OG) as substrates, the
latter generated by the tricarboxylic acid (TCA) cycle or by the transamination of
amino acids. This enzyme transfers the first oxygen atom to the proline residue, and
the second oxygen atom reacts with the 2-OD producing succinate. In addition to
oxygen and 2-OG, the PHD proteins require Fe2+ and ascorbate as cofactors [2].
The first 2-OG dioxygenase to be identified was procollagen prolyl hydroxylase, which participates in collagen synthesis and catalyzes trans-4-hydroxylation
reactions. There are three PHD isoenzymes: 1, 2, and 3. PHDs hydroxylate specific
proline residues inside a highly conserved domain whose sequence is LXXLAP
76
´
T.C. Ferreira, E.G.
Campos
Fig. 2 Mechanisms of HIF-1␣ regulation. Under normoxia conditions, HIF-1␣ is hydroxylated
by prolyl hydroxylase (PHD), an enzyme that belongs to the family of dioxygenases and requires
O2 and 2-oxoglutarate (2-OG) as substrates. This enzyme also uses iron as cosubstrate, which is
maintained in the reduced form (Fe+2 ) by ascorbic acid. Iron can be scavenged by treatment of
cells in normoxia with desferrioxamine (DFO), mimicking a hypoxia condition. Nickel and cobalt
also promote HIF-1␣ stabilization in normoxia condition by competing with iron. The pVHL recognizes the hydroxylated (OH) proline residues 402 and 564. pVHL is a member of E3-ubiquitinligase complex formed by elongin C (EloC), elongin B (EloB), and cullin 2 (CUL2) and RBX1.
Regulation of Glucose and Energy Metabolism in Cancer Cells by HIF-1
77
(where X indicates any amino acid; P indicates the hydroxylated proline; L, leucine
and A, alanine). Concerning the cellular localization, PHD1 is found exclusively
in the nucleus, PHD2 in the cytoplasm, and PHD3 is found in both cytoplasm
and nucleus, with predominance in the nucleus. Selective blockage of HIF-1␣
expression using RNA interference technology leads to a complete loss of PHD2
and PHD3 gene expression under hypoxia, whereas PHD1 mRNA levels remain
unchanged [2]. It seems that hypoxic induction of PHD2 and PHD3 is critically
dependent on HIF-1␣. This negative feedback mechanism probably limits HIF-1␣
accumulation in hypoxia and leads to accelerated degradation on reoxygenation
after long period of hypoxia. RNA interference showed the participation of PHD2 on
HIF-1␣ hydroxylation. Indeed, the specific silencing of PHD2 stabilized the HIF1␣ in normoxia conditions in several tested human cells. However, the silencing
of PHD1 and PHD3 has no effect in HIF-1␣ stabilization in both normoxia and
reoxgenation of cells briefly exposed to hypoxia. Therefore, it seems that PHD2
is the critical oxygen sensor that maintains HIF-1␣ in a low level in normoxic
conditions [3].
Under normoxia, HIF-1␣ activity is also regulated by hydroxylation of
asparagine 803 in the TAD-C domain, catalyzed by an enzyme known as factor
inhibiting HIF-1 (FIH), asparaginyl hydroxylase. FIH also belongs to the superfamily of 2-OD–dependent dioxygenases whose activity requires O2 as substrate.
The hydroxylation of Asn803 does not lead to HIF-1␣ degradation. Therefore, this
residue is not directly involved in the O2 -sensing mechanism. Hydroxylation of
Asn803 abolishes the interaction between HIF-1␣ and its transcriptional coactivators CBP and p300. CBP and p300 are coactivators that participate in the activities
of many different transcription factors. The CBP/p300 proteins have a key role in
coordinating and integrating multiple signal-dependent events. They make possible
the adequate expression levels of genes in response to diverse physiologic stimuli
and influences; for example, proliferation, differentiation, and apoptosis.
Fig. 2 (Continued) This complex is responsible for HIF-1␣ polyubiquitination prior to degradation by the proteasome. Factor inhibiting HIF-1 (FIH-1) hydroxylates the asparagine 803 (N803)
residue of HIF-1␣ preventing interaction with the CBP/p300. ARD1 acetylates the lysine 532
residue of HIF-1␣ leading to its degradation. Hypoxia, oncogenes, and growth factors can activate
the phosphatidylinositol 3-kinase (PI3K) pathway, which has AKT as a downstream target. AKT
can also be activated by phosphatidylinositol-dependent protein kinase 1 (PDK-1). PI3K/AKT
activation causes HIF-1␣ overexpression mediated by mTOR (mammalian target of rapamycin).
MEK (MAPK kinase) participates in the phosphorylation pathway activating the extracellular
signal–regulated kinases (ERK1/2), which phosphorylate HIF-1␣. Phosphorylated HIF-1␣ is stabilized and activated, enters into the nucleus, forms the heterodimer with HIF-1␤, is coactivated by
CBP/p300 complex, and induces the transcription of more than 100 target genes. A cysteine residue
is kept in the reduced state by redox factor-1 (REF-1) mediated by thioredoxin (TRX), which participate in HIF-1␣ stabilization. Under hypoxia, nitric oxide (NO) disfavors HIF-1␣ accumulation
(activates PHDs), and under normoxia condition, NO “mimics” hypoxia conditions (inhibits PHDs)
and leads to HIF-1␣ accumulation . Some evidence points to the role of reactive oxygen species
(ROS) as messengers that regulate HIF-1 activity (see Color Insert)
78
´
T.C. Ferreira, E.G.
Campos
pVHL is a component of the E3 ubiquitin-protein ligase complex that drives the
proteolysis in the proteasome, preventing the transcriptional activation of its target
genes. E3 ubiquitin ligases are a family of proteins that participate in the degradation and activity of many cellular proteins and function together with ubiquitinactivating enzyme E1 and ubiquitin-conjugating enzyme E2. In the hydroxylated
HIF-1␣ degradation pathway, E3 ubiquitin-protein ligase complex is formed by
pVHL, elongin C, elongin B, cullin 2, and RBX1 (Fig. 2). The E3 complex is
capable of functioning with E1 ubiquitin-activation and E2 ubiquitin-conjugating
enzymes to mediate the ubiquitination of HIF-1␣ leading to its rapid proteosomal
degradation. HIF-1␣ has a half-life of approximately 5 minutes. However, when
HIF-1␣ escapes the degradation pathway, its detection by Western blot experiments occurs within minutes. Indeed, HIF-1␣ is detectable in the nucleus of cells
after less than 2 minutes of exposure to anoxia or hypoxia. In this case, HIF-1␣
escapes ubiquitination and proteosomal degradation, is stabilized by posttranslational modifications, and enters into the nucleus. A comprehensive study about HIF1␣ turnover (analyzed by electrophoretic mobility shift assay; EMSA) showed that
its maximal protein expression occurs in cells exposed for 1 hour in anoxia/hypoxia.
These expression levels are maintained after 4 hours at low levels of oxygen exposure. However, after reoxygenation of cells, a decrease in the HIF-1 DNA binding
activity is observed after 2 minutes. In addition, the nuclear HIF-1␣ protein was
reduced within 4 to 8 minutes, down to a level below the detection limit within
32 minutes [4].
Under hypoxia conditions, HIF-1␣ is no longer degraded; it is phosphorylated
in the cytoplasm and activates its target genes inside the nucleus. Once the target genes are expressed, a physiologic response to low oxygen concentration is
achieved. Little is known about the fine-tuning of the posttranslational modifications
and their role in HIF-1␣ activity. The MAPK (mitogen-activated protein kinase)
and phosphatidylinositol-3-kinase (PI3K) pathways are certainly involved in the
regulation of HIF-1␣. ERK1/2 (also called p44/42) activation, which is a result
of activation of the upstream proteins Ras/Raf-1/MEK-1/ERK1/2, leads to HIF1␣ phosphorylation. Although MAPK phosphorylates HIF-1␣ in vitro, studies on
mutations in the phosphorylation sites in HIF-1␣ protein show that its transactivation capacity is not affected or directly mediated by phosphorylation. Regulation of
HIF-1␣ by MAPK is still an open issue [5]. Loss of tumor suppressor activity may
also contribute to HIF-1␣ overexpression. For example, interaction of the p53 tumor
suppressor gene with HIF-1␣ negatively regulates HIF-1␣ protein levels. Therefore, HIF-1␣ overexpression correlates with expression of mutant p53 in human
cancer.
AKT (protein kinase B), a serine/threonine kinase, is commonly activated
in cancer cells. This protein functions downstream to PIP3 (phosphatidylinositol 3,4,5-triphosphate) in the PI3K phosphorylation pathway (Fig. 2). PI3K/AKT
signaling pathway regulates vascular endothelial growth factor (VEGF) expression through HIF-1␣ in several types of transformed cells, including ovarian
cancer and human gastric adenocarcinoma. Cancer cells have constitutive AKT
Regulation of Glucose and Energy Metabolism in Cancer Cells by HIF-1
79
activity through amplification of PI3K signaling pathway or deletion of phosphatase and tensin homolog deleted on chromosome 10 (PTEN), a PI3K antagonist. AKT induction is an important event in tumorigenesis as this pathway leads
to a higher level of glycolysis in a dose-dependent manner that correlates with
a more aggressive malignancy in vivo. Apparently, AKT has an effect more pronounced on glucose metabolism than it effects on proliferation and survival of
these transformed cells. This phenomenon was observed in established human
glioblastoma cells, which showed a switch to aerobic glycolysis and glucose
dependence [6].
HIF-1␣ is also modified in the lysine residue 532 localized in the ODD domain
of HIF-1␣ by an acetyl transferase called arrest defective-1 (ARD1). ARD1 was
originally identified in the yeast Saccharomyces cerevisiae and received this name
due to the phenotype of mutant yeast defective in the mitotic cell cycle. A human
variant of ARD1 is able to partially decrease HIF-1␣ accumulation in hypoxia
[7]. Another HIF-1␣ posttranslational modification is S-nitrosation. Nitric oxide
(NO) promotes HIF-1␣ stabilization, DNA binding, and transactivation of target
genes under normoxic conditions. This NO effect on HIF-1␣ activity was verified
in several cell lines, including swine tubular cells (LLC-PK1), human glioblastoma
and hepatoma, as well as endothelial cells from bovine pulmonary artery. These
cells, when treated with different NO donors, such as S-nitroglutathione, (Z)-1[2-(2-aminoethyl)-N-(2-ammonioethyl)amino]diazen-1-ium-1,2-diolate (NOC-18),
1-hydroxy-2-oxo-3-(3-aminopropyl)-3-isopropyl-1-triazene (NOC-5), and S-nitroso
-N-acetyl-D,L-penicillamine (SNAP), accumulate HIF-1␣ protein [8]. Therefore,
under hypoxia, NO disfavors HIF-1␣ accumulation, whereas under normoxic
conditions, NO “mimics” hypoxia conditions and leads to HIF-1␣ accumulation.
4 HIF-1 Isoforms
There are two other members of the bHLH-PAS superfamily that share structural similarity to HIF-1␣: HIF-2␣, also named endothelial PAS-domain protein
1 (EPAS1), and HIF-3␣. The roles of the 2␣ and 3␣ subunits are poorly understood. Similarly to HIF-1␣, HIF-2␣ is also regulated by hydroxylation of the conserved proline residue, which causes its degradation under normoxic conditions via
ubiquitin-E3-ligase. HIF-2␣ has 48% amino acid sequence identity with HIF-1␣.
HIF-2␣ and HIF-3␣ also have the common ability to heterodimerize to HIF-1␤
and bind to the hypoxia response element (HRE) motif in the regulatory regions of
target genes. Additionally, an alternative splice variant of HIF-3␣ exists, a protein
that lacks the transactivation domain. This splice variant was termed inhibitory PAS
(IPAS) protein, and it was identified as the dominant-negative regulator of HIF-1.
IPAS interacts with the amino-terminal region of HIF-1␣ and prevents its DNA
binding. The effect of the different members of the HIF-1 family on gene expression can vary depending on the cell type. For example, HIF-1␣ and HIF-2␣ are
strongly expressed in the kidney, vascular cells, bone marrow, macrophages, lungs,
80
´
T.C. Ferreira, E.G.
Campos
endothelium, and carotid body, whereas HIF-3␣ is mainly expressed in Purkinje
cells of the cerebellum and mouse cornea epithelial cells. HIF-2␣ has more restrictive distribution inside tissues than does HIF-1 [9].
5 HIF-1␣ Induction by Nonhypoxic Stimuli
In addition to hypoxia, HIF-1 is also responsive to several nonhypoxic stimuli, including insulin, platelet-derived growth factor (PDGF), transforming growth
factor ␤ (TGF-␤), insulin-like growth factor (IGF-1), thrombin angiotensin
II, cytokines, and carbachol (activates the muscarinic acetylcholine receptors).
Whereas growth factor stimulation of HIF-1␣ is cell type specific, hypoxia increases
HIF-1␣ expression in all cell types. In addition, HIF-1 is activated by some metals
such as cobalt, chromium, nickel, arsenic, and iron. Desferrioxamine also activates
HIF-1 by mimicking hypoxic conditions. Some evidences point to the role of reactive oxygen species (ROS) as messengers that regulate HIF-1 activity [10].
The oxidative status of the iron ion bound to the PHDs might also regulate
HIF-1␣ activity, as these enzymes are Fe2+ -dependent. ROS generated by mitochondria and released into the cytoplasm of cells exposed to hypoxia promote oxidation
of Fe2+ to Fe3+ . A decrease in the Fe2+ availability would cause reduction in the
PHD activity. This mechanism allows HIF-1 accumulation and stabilization [11].
6 HIF-1 in Cancer
Once stabilized and activated, inside the nucleus, HIF-1 recognizes and binds to
hypoxia response elements (HREs) present in the regulatory regions of its target genes. The consensus HRE has the sequence 5 -BRCGTGVBBB-3 , where the
RCGTG core is the HIF-1 binding site (HBS). Microarray experiments suggest that
more than 200 HIF-1 target genes might exist. However, not all of them are likely to
be directly regulated by an HRE in their regulatory regions [10]. More than 100 HIF1 target genes are known, including the genes that encode erythropoietin, VEGF,
glucose transporters, and glycolytic enzymes.
Erythropoietin (EPO) and VEGF play crucial roles in adaptive responses to systemic and local hypoxia. Both are implicated in tumor progression. EPO stimulates
erythropoiesis, which increases O2 delivery, whereas VEGF stimulates angiogenesis, which increases delivery of both O2 and energy substrates such as glucose. Rigorous analysis has confirmed the role of HIF-1 in EPO and VEGF transcriptional
activation, and consequently in carcinogenesis.
HIF-1␣ is overexpressed in many human cancers [12]. Immunohistochemical
staining analysis of surgical biopsy specimens show overexpression of HIF-1␣ protein in colon, ovary, lung, prostate, skin, stomach, brain, and breast cancers. Significant associations between HIF-1␣ overexpression and patient mortality have been
shown in cancers of the breast, brain, cervix, oropharynx, esophagus, ovary, and
Regulation of Glucose and Energy Metabolism in Cancer Cells by HIF-1
81
uterus. Premalignancy and neomalignant lesions as well as cancer metastases of
some cancer types also show increased HIF-1␣ protein. For example, HIF-1␣ is
overexpressed in the majority of breast and colon cancer metastases. By contrast,
association between HIF-1␣ overexpression and decreased mortality were reported
for patients with head and neck cancer and non–small cell lung cancer. Thus, the
effect of HIF-1␣ overexpression seems to be dependent on the cancer type. HIF-1␣
can have antiapoptotic or proapoptotic effects on tumor cells [9].
Mutation in the VHL gene results in the so called von Hippel–Lindau syndrome,
characterized by the formation of tumors and fluid-filled sacs (cysts) in many different parts of the body. Clear cell renal carcinoma (CCRC) cells lacking pVHL
tumor suppressor constitutively express HIF-1␣ and HIF-1 target genes under nonhypoxic conditions [13]. CCRCs present all the main characteristics supposed to be
maintained by proteins encoded by genes regulated by HIF-1: increased glycolysis,
decreased respiration, increased glucose uptake and lactate production. pVHL lossof-function also occurs in hemangioblastoma. Therefore, at least in these two types
of tumors, deregulation of HIF-1␣ regulatory pathway plays a causal role.
7 HIF-1, Cancer, and the Glycolytic Pathway
The way cells change their metabolism under hypoxia shares many similarities to
the way tumors cells change their metabolism to survive and proliferate. Indeed,
many of the known oncogenic signaling pathways overlap with hypoxia-induced
pathways. Cancer cells have special metabolic needs for their proliferation and survival. They need to cope with different levels of oxygen and adjust their metabolism
to guarantee cell proliferation and minimize apoptosis. The vast majority of human
and animal tumors display a high rate of glycolysis, increased glucose uptake,
increased lactate production, and decreased respiration under aerobic conditions
(a phenomenon known as the Warburg effect) [14]. Aerobic glycolysis provides
cancer cells (both nonmetastatic and metastastic) a high competitive capacity over
normal cells. This adaptive response can be driven by hypoxia, as well as by mutations that inactivate tumor suppressor genes or activate oncogenes. Hypoxia is a
potent inducer of gene expression, especially of genes involved in glycolysis for
maintaining cellular energy. The higher glycolytic rate in tumor cells compared
with that of nontumorigenic cells can be attained by the increased transcription,
followed by translation, of glycolytic genes and glucose transporters associated or
not with inhibition of oxidative phosphorylation. Hypoxia has HIF-1 as the key
regulator. Evidence is surmounting that HIF-1 regulation of glycolysis is important in carcinogenesis. However, there is much controversy as to whether HIF-1–
induced glycolysis is a reactive process or a “cause” of cancer. A definitive solution
to this key question is complicated by the fact that hypoxia and HIF-1 are not exclusive regulators of the glycolytic enzymes, and HIF-1 activation is not associated
only with hypoxia (oncogenes and growth factors can also activate this pathway).
In addition, high pyruvate concentrations result in HIF-1␣ stabilization indepen-
82
´
T.C. Ferreira, E.G.
Campos
dently of hypoxia [15].Therefore, although hypoxic HIF-1 activation alone does
not explain aerobic glycolysis, other factors may provide feedback to enhance normoxic HIF-1 induction. The structure, regulation, and action of HIF-1 was discussed
in detail above and shows a very complex network (Fig. 2). Regardless of which
pathway leads to HIF-1 activation, its effect on glucose metabolism is due to the
upregulation of specific target genes.
Regulation of metabolism by HIF-1␣ was evidenced through investigation of its
function in mouse embryonic stem (ES) cells in which the Hif1␣ gene was inactivated by homologous recombination. Analysis of Hif1a+/+ , Hif1a+/- , and Hif1a-/ES cells revealed that HIF-1␣ is required for the induction of several genes whose
products are involved in glucose metabolism; more specifically, aldolase A and C,
enolase 1, hexokinase I and II, phosphofructokinase L, phosphoglycerate kinase I,
pyruvate kinase M, glyceraldehyde-3-phosphate dehydrogenase, lactate dehydrogenase A, GLUT1 and GLUT3 [16]. HRE has been characterized in the aldolase A,
enolase 1, and lactate dehydrogenase A human promoters and also in the aldolase C
mouse promoter. Complete identification of the key gene products specifying HIFdependent tumor metabolism is crucial for therapeutic targeting of HIF-1 signaling
in cancer.
In the presence of oxygen, glucose is converted to pyruvate by the glycolytic
enzymes. Pyruvate is transported to the mitochondrial matrix, where it is converted
to acetyl-CoA and metabolized in the TCA cycle. Under hypoxia, pyruvate is converted to lactate through lactate dehydrogenase (LDH) A in the cytoplasm, which
is a target for HIF-1 regulation. In cancer cells, aerobic glycolysis and lactate production is favored to the detriment of oxidative phosphorylation. To understand the
role of HIF-1 in the regulation of glucose and energy metabolism in cancer cells,
we reviewed the literature in search of experimental data demonstrating the direct
involvement of HIF-1 in the induction of such genes in these cells. In some cases,
the link between HIF-1 and upregulation of a gene in cancer cells was based on the
presence of HRE motifs in their regulatory regions. In the next sections, we describe
the link found between HIF-1 and the proteins involved in glucose metabolism in
cancer cells.
8 Hexokinase
Hexokinases (HKs) catalyze the first step in the glycolytic pathway, conversion
of glucose to glucose-6-phosphate (G6P). Human cells express four isozymes of
hexokinases: types I, II, III, and IV (glucokinase). The key hexokinase isoenzyme
involved in cancer promotion is the type II (HKII), the major isozyme overexpressed
in tumors exhibiting the high glycolytic phenotype. HKII has been extensively studied by Dr. Peter Pedersen from Johns Hopkins University School of Medicine and
his co-workers [17]. They found that HKII binds to transmembrane channels formed
by the porin-like protein voltage-dependent anion channel (VDAC) located within
the outer mitochondrial membrane. This interaction reduces the enzyme’s sensitivity
Regulation of Glucose and Energy Metabolism in Cancer Cells by HIF-1
83
to product inhibition by G6P and allows the enzyme to have direct and rapid access
to mitochondrially generated ATP. It also protects HKII against proteolytic degradation. These combined properties, together with the high content of the enzyme
in highly malignant tumors (>100-fold increase), result in the rapid production of
G6P. This key metabolic intermediate-precursor serves as a major carbon source
for most biosynthetic pathways that are essential for the growth and rapid proliferation of tumors. It also serves as the initial substrate for glycolysis and ATP
generation.
Enhanced transcription of HKII occurs in malignant tumors, implicating the
HKII promoter activation and upregulation during tumorigenesis. Functional HREs
(HIF-1 responsive) are located in the distal region of the HKII promoter. HKII can
also be transcriptionally activated by mutant p53 or through demethylation of its
promoter. Regulation of the gene encoding hepatoma HKII occurs through activation of its promoter that contains in its distal region two overlapping E-box/HIF-1
sequences separated by approximately 50 base pairs. These sequences are known
to be associated in other gene promoters with responsiveness to glucose (E-box) or
hypoxic conditions (HIF-1). The response of the HKII gene promoter to both glucose and hypoxic conditions is not confined to its distal region but involves significantly the proximal region as well. Elements in this region remain to be identified
and studied in detail.
9 Glucose Phosphate Isomerase
Glucose phosphate isomerase (GPI) is a housekeeping cytosolic enzyme of glucose
metabolism that plays a key role in both glycolysis and gluconeogenesis pathways, catalyzing the reversible isomerization of glucose-6-phosphate and fructose6-phosphate. Upon secretion, GPI also acts as a cytokine also referred to as autocrine
mobility factor (AMF) or NLK [18]. AMF was originally isolated as a cytokine
autocrine factor that stimulates cell mobility and NLK as a lymphokine and neurotrophic factor for spinal and sensory neurons. GPI/AMF/NLK is associated with
the development of tumors, and an increased amount of this protein is present in the
urine and serum of patients with various malignancies. Therefore, in addition to its
role in the glycolytic pathway, the induction of this gene under hypoxia may play a
role in stimulating cell mobility, invasion, and metastasis in tumor cells.
In human pancreatic cancer PC-3 cells, GPI mRNA is weakly expressed in cells
cultured in normoxic conditions, but its synthesis is markedly increased in hypoxic
conditions (1% oxygen plus CoCl2 ). YC-1, 3-[5 -hydroxymethyl-2 furyl]-1-benzyl
indazole, a guanylate cyclase activator, reduces the mRNA levels of GPI in PC-3
cells in a dose-dependent manner under hypoxic conditions. Guanylate cyclase activation can increase production of NO, which inhibits the accumulation and activity
of HIF-1 in hypoxia. Therefore, GPI transcriptional activation is intimately linked
to HIF-1 activity.
84
´
T.C. Ferreira, E.G.
Campos
10 Phosphofructokinase
Phosphofructokinase 1 (PFK-1) catalyses the rate-limiting phosphorylation of
fructose-6-phosphate to fructose-1,6-biphosphate, which is an energy-consuming
step in glycolysis. The fructose-2,6-biphosphate is the most potent allosteric
activator of glycolysis and exerts control over the rate of glucose utilization. It
activates PFK-1 and inhibits the gluconeogenic enzyme fructose-1,6-biphosphatase.
Because of the antagonistic effects in these enzymes, the product of the reaction
catalyzed by 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase (PFKBF) plays
an essential role in the opposing glycolytic and gluconeogenic pathways.
In human cells, there are at least four different genes encoding PFKFBs
(PFKBP-1, PFKBP-2, PFKBP-3, and PFKBP-4). They all respond to hypoxia in
vivo, and this response is diverse in different organs. Tissue-specific isoforms of
PFKBP are not completely exclusive, and several tissues express more than one
isoform. Among the PFKFB isoforms of mammalian origin, PFKFB3 has the highest kinase:phosphatase activity (K/B) and is the most highly expressed isozyme in
transformed cells. The PFKFB-4 isozyme lacks a serine phosphorylation residue
that is critical for the downregulation of its kinase activity. Therefore, one can conclude that it may have a high K/B ratio, and it should greatly promote glycolysis
under conditions of limited oxygen supply. The use of dimethyloxalylglycine (a
specific competitive inhibitor of PHDs) demonstrated the role of HIF-1␣ in hypoxic
induction of the genes coding for PFKFB3 and PFKFB-4 isoforms (in human hepatoma Hep-3B and HepG2 cell lines, human prostate cancer cell, PC-3 and HeLa
cells). A splice variant of PFKFB-4 (PFKFB-4 s) is also induced by hypoxia in DB-1
melanoma, besides the full-length PFKFB-4. Hypoxic induction of the transcription
of these genes is mediated by HREs located in their promoter regions. In addition,
PFKFB3 and PFKFB4 isoforms are overexpressed in solid malignant tumors from
the breast and colon, overexpression of the PFKFB-4 protein being higher compared
with that of PFKBF-3 [19].
11 Aldolase
Aldolase (ALD) catalyzes the reversible aldol cleavage of fructose-6-phosphate into
two trioses, glyceraldehyde-3-phosphate and dihydroxyacetone. It has a role in the
glycolytic, gluconeogenic, and fructose metabolic pathways. Three tissue-specific
forms exist in higher vertebrates: aldolase A (predominately in muscle and red blood
cells), aldolase B (predominately in liver, kidney, and intestine), and aldolase C (predominately in brain). The direct association of aldolase gene expression and HIF-1
in tumors was observed in a study with a mouse hepatoma cell line, Hepa1c1c7,
and its mutant derivative, ARNT-deficient c4T. The absence of the ARNT protein
completely abolished the oxygen regulation of the aldolase A and phosphoglycerate kinase 1 (PGK1) gene expression, indicating that these genes share a common
pathway of response to hypoxia involving activation of HIF-1. Among the three
Regulation of Glucose and Energy Metabolism in Cancer Cells by HIF-1
85
isoenzymes, aldolases A and C seem to have evolved to perform the glycolytic
reaction, fructose-1,6-phosphate cleavage, more efficiently than does aldolase B.
Aldolase B seems to be more involved with the gluconeogenic pathway [20].
12 Glyceraldehyde Phosphate Dehydrogenase
Glyceraldehyde phosphate dehydrogenase (GAPDH) is a key enzyme in glycolysis
and also a multifunctional protein. It converts glyceraldehyde-3-phosphate into 1,3bisphosphoglycerate. Its gene is constitutively expressed in many tissues. Variations
in GAPDH expression levels occurs in tissues at different developmental stages and
have been observed in cells treated with a variety of agents such as mitogens, dexamethasone, calcium, insulin, and in cells under hypoxia conditions. Overexpression
of GAPDH in tumor cells as a consequence of the development of a hypoxic cellular microenvironment is cell type specific. It has been observed in human prostate
adenocarcinoma cells (cell line LNCap), human spontaneous cervical cancer cells
(SiHA), but not in human malignant glioma cells (cell lines U373-MG, GaMG, and
U251) ([21], [22]). It is not clear if the reason for the upregulation of GAPDH in
tumor cells is only for increasing glycolysis. The role of GAPDH as a NADH generator and as a mediator of cell death has been highlighted in several studies, which
adds to the complexity of its upregulation in tumor cells.
13 Phosphoglycerate Kinase
Phosphoglycerate kinase (PGK) catalyzes the transfer of the phosphoryl group
from the acyl phosphate of 1,3-bisphosphoglycerate to ADP. ATP and 3phosphoglycerate are the products. Human PGK1 gene harbors HREs in the 5
flanking and 5 UT regions, and mouse PGK1 harbors three HREs in the 5 flanking sequence [10]. The HREs of the PGK promoter mediate the transcriptional
response to hypoxia. Indeed, the induction of PGK in different solid tumor cell lines
in hypoxic conditions is under the control of HIF-1␣. Expression of PGK1 increases
in mouse hepatoma Hepa1c1c7 cell lines exposed to hypoxia or to treatment with
desferrioxamine [20].
14 Enolase
Enolase (ENO) is also upregulated in many cancers, including breast, liver, colon,
and lung. Direct evidence of HIF-1 upregulation of the enolase gene was obtained
using YC-1, a drug that targets HIF-1. Levels of aldolase and enolase mRNAs
decrease in YC-1–treated Hep3B hepatoma cells. The inhibitory action of YC-1
on these glycolytic genes probably inhibits cell survival under hypoxia and may
promote cell death in hypoxic areas [23].
86
´
T.C. Ferreira, E.G.
Campos
15 Pyruvate Kinase
Pyruvate kinase (PK) is responsible for net ATP production within the glycolytic
pathway. It is consistently altered during tumorigenesis. Four isoenzymes of pyruvate kinase have been identified: type L (present in tissues with gluconeogenesis
such as liver and kidney); type R (expressed in erythrocytes); type M1 (muscle
and brain); and type M2 (expressed in lung tissues and all cells with high rates
of nucleic acid synthesis, such as embryonic cells, adult stem cells, and especially
tumor cells). Investigation of the mRNA levels of HIF-1␣, PGK1 and muscle-type
pyruvate kinase 2 (PKM2) were performed in various hepatoma cell lines as well
as in mouse and human tumor specimens and the corresponding normal tissues. In
five of eight human colorectal cancers investigated, PGK1 and PKM2 mRNA levels were increased in comparison with the corresponding normal tissues, whereas
HIF1-␣ was not significantly changed [24]. Though probable, these studies do not
conclusively demonstrate that the increase in PKM2 gene expression is, in fact,
mediated by HIF-1–induced transcriptional activation. However, HIF-1␣ plays a
critical role in PK expression during hypoxic challenge in glial cells.
16 Pyruvate Dehydrogenase
The pyruvate dehydrogenase (PDH) complex catalyzes the conversion of pyruvate
to acetyl-coenzyme A, which enters the TCA cycle, producing NADH and carbon
dioxide. The PDH complex is composed of three subunits, E1 (pyruvate decarboxylase), E2 (dihydrolipoamide acetyltransferase), and E3 (dihydrolipoamide dehydrogenase), and is under the control of pyruvate dehydrogenase kinases (PDKs)
1 to 4. PDKs phosphorylate the E1 subunit of PDH causing its inhibition. The
activity of PDK is regulated by the concentration of the metabolic products of
pyruvate (NADH and acetyl-coA). HIF-1 upregulates the expression of PDK1
[25]. Upregulation of the expression of PDK1 by HIF-1 illustrates how it favors
a metabolic switch toward glycolysis to the detriment of oxidative phosphorylation. Hypoxic cells compensate the inefficient glycolytic energy output by increasing glucose uptake, overexpressing and increasing the activities of the glycolytic
enzymes.
17 Lactate Dehydrogenase and Lactate Carrier MCT4
Lactate dehydrogenase (LDH) converts pyruvate into lactate. LDH-A converts pyruvate to lactate during glycolysis under anaerobic conditions in normal cells. The
other isoenzyme, LDH-B, kinetically favors the conversion of lactate to pyruvate
and is found at high levels in aerobic tissues such as heart. Elevated LDH-A expression contributes to the ability of tumor cells to undergo aerobic glycolysis. Once
produced, lactic acid exits cancer cells through the MCT4 lactate carrier that is also
Regulation of Glucose and Energy Metabolism in Cancer Cells by HIF-1
87
induced in these cells. Under hypoxia, both pyruvate transformation into lactate
with concomitant regeneration of NAD+ and lactate exit from the cell are stimulated
because the genes encoding the LDH-A isoform and the lactate carrier MCT4 are
HIF-1 targets.
The c-myc gene is frequently activated in human cancers. This transcription factor forms a heterodimer with another protein (MAX) and binds to a
5 -CACGTG-3 consensus sequence in its target genes inducing gene expression. In
addition to HIF-1, the LDH-A gene is also a c-MYC target. Therefore, the molecular
basis for altered tumor glycolysis and lactate production cannot be solely explained
by HIF-1 activation of specific genes [26]. In addition to its role in lactate production, LDH-A may also be involved in the regulation of gene expression and/or
DNA replication. Therefore, its modulation by HIF-1 and c-MYC could be linked
to effects other than to increased lactate production.
18 Glucose Transporters and HIF-1
Glucose enters animal cells through facilitated diffusion via glucose transporters
(GLUT1 to GLUT5) and is subsequently phosphorylated by hexokinase. GLUT1
and GLUT3, present in nearly all mammalian cells, are responsible for basal glucose uptake and continually transport glucose into cells at an essentially constant rate. They have a high affinity for glucose, allowing transport of glucose at
a high rate under normal physiologic conditions. Cancer cells overexpress both
GLUT1 and GLUT3 to increase glucose uptake. Upregulation of glucose transporters has been observed during carcinogenesis in esophageal, gastric, breast,
and colon cancers. A major technique used to measure glucose intake in tumor
cells is positron emission tomography (PET) with [18 F]fluoro-2-deoxy-D-glucose
(FDG). PET is a noninvasive imaging technique used in the clinic to detect malignant tumors, which measures the glucose metabolism in vivo. FDG is a fluorinated glucose analogue and accumulates in tumor cells in proportion to the rate
of glucose metabolism. A high uptake of FDG is associated with a higher GLUT1
expression.
The enhanced uptake of glucose often reflects tumor aggressiveness in patients
with different types of tumors. The glucose transfer from the bloodstream to the
inside of cells mediated by GLUT1 plays a central role in the development and
malignant behavior of cancer cells. HIF-1 binding to the promoter of glucose transporters was seen in human HepG2 cells (GLUT1 and GLUT3) and mouse hepatoma
cells exposed to hypoxia (GLUT1). Studies of GLUT mRNA expression in hypoxic
cells have shown that the level of GLUT1 mRNA in human placenta choriocarcinoma cell line BeWo (JCRB911) increases 1.4-fold in the presence of 5% O2 and
by 2.2-fold in the presence of 250 ␮M CoCl2 compared with that under 20% O2 .
In the rat choriocarcinoma Rcho-1 cell line, the level of GLUT1 mRNA increases
1.8-fold in the presence of 5% O2 and by 3.2-fold in the presence of 250 μ M CoCl2
compared with that under 20% O2 . The GLUT1 protein levels also increase under
88
´
T.C. Ferreira, E.G.
Campos
hypoxia and CoCl2 treatment, and these results suggest that GLUT mRNA and protein levels were directly mediated through HIF-1 induction [27].
19 HIF-1␣ Upregulation by TCA Intermediates Is Linked
to Carcinogenesis
Succinate dehydrogenase (SDH) is a TCA enzyme, which is also part of the
mitochondrial respiratory chain. Succinate dehydrogenase deficiency causes either
severe encephalomyopathy or tumor formation [28]. Germ-line heterozygote mutations in the SDH subunits cause inherited pheochromocytomas (adrenal gland
tumors) and paragangliomas. The effects of SDH deficiency are postulated to be
associated with the disturbance of two signaling pathways: one involving the superoxide radical and another involving HIF-1. SDH defect can increase superoxide radical production by the mitochondrial respiratory chain, which is able to trigger both
cell death and proliferation. On the other hand, when succinate accumulates inside
the mitochondria, it is transported into the cytosol where it can inhibit prolyl hydroxylases (PHD1 to PHD3), the enzymes responsible for HIF-1 destabilization. Upon
PHD inhibition, HIF-1␣ is stabilized under normoxic conditions. Heterodimerization of stabilized HIF-1␣ and HIF-1␤ forms active HIF-1, which induces transcription of nuclear genes involved in tumor progression.
Mutations in another TCA enzyme, fumarate hydratase (FH), cause cutaneous
and uterine leiomyomas, as well as renal cell carcinomas. Mitochondrial proteins
may also play a role in the development of sporadic kidney tumor oncocytoma. High
concentrations of cytosolic fumarate may inhibit prolyl hydroxylases and have the
same consequences as excess succinate on HIF-1 destabilization. Defective SDH
and FH activity lead, respectively, to the accumulation of succinate and fumarate.
These TCA intermediates, in addition to oxaloacetate, are capable of inhibiting the
PHD activity and subsequently inhibiting HIF-1␣ degradation [29].
20 HIF-1 Activity as a Therapeutic Target
The prognostic significance of HIF-1␣ expression in cancer cells, its central role in
the adaptive response to hypoxia, and its participation in the Warburg effect make
it a potential target for anticancer drug development. It is a reasonable speculation that inhibition of HIF-1 activity may be a useful cancer therapy. A number of
anticancer agents that may target HIF-1 activity have been described (some cited
in this text) and are treated more fully by Patiar and Harris [30]. The challenge
remains to prove that the antitumorigenic effect of these drugs is due to HIF-1␣
inhibition. Selectivity may be resolved by the fact that tumor cells are more hypoxic
than are normal cells and have specific intrinsic markers of HIF-1␣ activation. Further increase in this difference is possible through the use of angiogenesis inhibitors
(e.g., thrombospondin-1) in association with HIF-1␣ inhibition.
Regulation of Glucose and Energy Metabolism in Cancer Cells by HIF-1
89
21 Conclusion
HIF-1␣ can be activated by hypoxic and nonhypoxic stimuli, including growth factors, oncogenes, metals, and ROS. Metabolism of glucose and oxidative phosphorylation are directly affected by the HIF-1 signaling pathway. A change in ATP
production pathway from oxidative phosphorylation to aerobic glycolysis is a hallmark of cancer cells. There is an evident overlap between HIF-1 function and cancer
cell metabolism. Cancer cells have special needs for rapid proliferation and survival.
Some of these needs are fulfilled by the HIF-1 signaling cascade. In addition, HIF-1
is located at the center of the major pathways that define cell destiny toward adaptation to fast growth or apoptosis. The available data show that cancer cell survival
is a much more pronounced HIF-1 effect than is cell death. Therefore, HIF-1 will
continue to be explored as a potential target for cancer therapy.
References
1. Semenza GL. Expression of hypoxia-inducible factor 1: mechanisms and consequences.
Biochem Pharmacol 2000;59: 47–53.
2. Metzen E, Berchner-Pfannschmidt U, Stengel P, et al. Intracellular localisation of human
HIF-1␣ hydroxylases: implications for oxygen sensing. J Cell Sci 2003; 116:1319–1326.
3. Berra E, Benizri E, Ginouv`es A, Volmat V, Roux D, Pouyss´egur J. HIF prolyl-hydroxylase
2 is the key oxygen sensor setting low steady-state levels of HIF-1␣ in normoxia. EMBO J
2003; 22:4082–4090.
4. Jewell UR, Kvietikova I, Scheid A, Bauer C, Wenger RH, Gassmann M. Induction of HIF–1␣
in response to hypoxia is instantaneous. FASEB J 2001; 15:1312–1314.
5. H¨opfl G, Ogunshola O, Gassmann M. HIFs and tumors – causes and consequences. Am J
Physiol Regul Integr Comp Physiol 2004; 286:R608–R623.
6. Elstrom RL, Bauer DE, Buzzai M, et al. AKT stimulates aerobic glycolysis in cancer cells.
Cancer Res 2004; 64:3892–3899.
7. Kim SH, Park JA, Kim JH, et al. Characterization of ARD1 variants in mammalian cells.
Biochem Biophys Res Commun 2006; 340:422–427.
8. Br¨une B, Zhou J. The role of nitric oxide (NO) is stability regulation of hypoxia inducible
factor-1␣ (HIF-1␣). Current Med Chem 2003; 10:845–855.
9. Lee J-W, Bae S-H, Jeong J-W, Kim S-H, Kim K-W. Hypoxia-inducible factor (HIF-1) ␣: its
protein stabilization and biological functions. Exp Mol Med 2004; 36:1–12.
10. Wenger RH, Stiehl DP, Camenisch G. Integration of oxygen signalling at the consensus HRE.
Sci Stke 2005;re12.
11. Pouyssegur J, Mechta-Grigoriou F. Redox regulation of the hypoxia-inducible factor. Biol
Chem 2006; 387:1337–1346.
12. Zhong H, De Marzo AM, Laughner E, et al. Overexpression of hypoxia-inducible factor
1alpha in common human cancers and their metastases. Cancer Res 1999; 59:5830–5835.
13. Semenza GL. HIF-1 mediates the Warburg effect in clear cell renal carcinoma. J Bioenerg
Biomembr 2007; 39:231–234.
14. Warburg O. On the origin of cancer cells. Science 1956; 123:309–314.
15. Lu H, Forbes RA, Verma A. Hypoxia-inducible factor 1 activation by aerobic glycolysis implicates the Warburg effect in carcinogenesis. J Biol Chem 2002; 277:23111–23115.
16. Iyer NV, Kotch LE, Agani F, et al. Cellular and developmental control of O2 homeostasis by
hypoxia-inducible factor 1␣. Genes Dev 1998; 12:149–162.
90
´
T.C. Ferreira, E.G.
Campos
17. Mathupala SP, Ko YH, Pedersen PL. Hexokinase II: cancer’s double-edged sword acting as
both facilitator and gatekeeper of malignancy when bound to mitochondria. Oncogene 2006;
25:4777–4786.
18. Funasaka T, Yanagawa T, Hogan V, Raz A. Regulation of phosphoglucose isomerase/autocrine
motility factor expression by hypoxia. FASEB J 2005; 19:1422–1430.
19. Minchenko OH, Ochiai A, Opentanova IL, et al. Overexpression of 6-phosphofructo2-kinase/fructose-2,6-bisphosphatase-4 in the human breast and colon malignant tumors.
Biochimie 2005; 87:1005–1010.
20. Salceda S, Beck I, Caro J. Absolute requirement of aryl hydrocarbon receptor nuclear translocator protein for gene activation by hypoxia. Arch Biochem Biophys 1996; 334:389–394.
21. Lu S, Gu X, Hoestje S, Epner DE. Identification of an additional hypoxia responsive element in the glyceraldehyde-3-phosphate dehydrogenase gene promoter. Biochim Biophys
Acta 2002; 1574:152–156.
22. Said HM, Hagemann C, Stojic J, et al. GAPDH is not regulated in human glioblastoma under
hypoxic conditions. BMC Mol Biol 2007; 8:55.
23. Yeo EJ, Chun YS, Cho YS, et al. YC-1: a potential anticancer drug targeting hypoxiainducible factor 1. J Natl Cancer Inst 2003; 95:516–525.
24. Kress S, Stein A, Maurer P, et al. Expression of hypoxia-inducible genes in tumor cells.
J Cancer Res Clin Oncol 1998; 124:315–320.
25. Kim JW, Tchernyshyov I, Semenza GL, Dang CV. HIF-1-mediated expression of pyruvate
dehydrogenase kinase: a metabolic switch required for cellular adaptation to hypoxia. Cell
Metab 2006; 3:177–185.
26. Shim H, Dolde C, Lewis BC, et al. c-Myc transactivation of LDH-A: implications for tumor
metabolism and growth. Proc Natl Acad Sci USA 1997; 94:6658–6663.
27. Hayashi M, Sakata M, Takeda T, et al. Induction of glucose transporter 1 expression through
hypoxia-inducible factor 1 under hypoxic conditions in trophoblast-derived cells. J Endocrinol
2004; 183:145–154.
28. Selak MA, Armour SM, MacKenzie ED, et al. Succinate links TCA cycle dysfunction to
oncogenesis by inhibiting HIF-alpha prolyl hydroxylase. Cancer Cell 2005; 7:77–85.
29. Pollard PJ, Briere JJ, Alam NA, et al. Accumulation of Krebs cycle intermediates and overexpression of HIF1alpha in tumours which result from germline FH and SDH mutations. Hum
Mol Genet 2005; 14:2231–2239.
30. Patiar S, Harris AL. Role of hypoxia-inducible factor-1alpha as a cancer therapy target.
Endocr Relat Cancer 2006; 13(Suppl 1):S61–S75.
The Role of Glycolysis in Cellular
Immortalization
Hiroshi Kondoh
Abstract The clinical significance of the Warburg effect is well established,
although it is not clear why and when cancer cells start to display a high glycolytic
rate in vivo. The discovery of hypoxia-inducible transcriptional factor as a critical
regulatory factor for glycolytic genes along with the recent advances in senescence
biology imply that the metabolic shift to enhanced glycolysis would be observed in
multiple stages during tumorigenesis in vivo. Increased glycolysis would be required
in the early step of immortalization to avoid oxidative stress–mediated senescence,
followed by the adaptation to hypoxic conditions through increased enhancement
of glycolysis during transformation. Discovery of regulatory mechanisms for such
a metabolic shift could provide important information for the future development of
cancer diagnosis and anticancer therapy.
Keywords Phosphoglycerate mutase · Glycolysis · Immortalization · p53 ·
Warburg effect
1 In Which Stage of Carcinogenesis Is the Warburg Effect
Required?
1.1 Immortalization and Transformation During Multistep
Carcinogenesis
It is well-known that tumorigenesis is a multistep process that involves a series
of genetic and epigenetic alterations [1]. Chemically induced skin tumor is the
most established animal model of multistep tumorigenesis. The sequential application of carcinogen can effectively induce tumors in mice. The application of 7,
H. Kondoh (B)
Department of Geriatric Medicine, Graduate School of Medicine, Kyoto University,
54 Kawahara-cho, Shogoin, Sakyo-ku, Kyoto, 606-8507, Japan
e-mail: hkondoh@kuhp.kyoto-u.ac.jp
S.P. Apte, R. Sarangarajan (eds.), Cellular Respiration and Carcinogenesis,
DOI 10.1007/978-1-59745-435-3 7,
C Humana Press, a part of Springer Science+Business Media, LLC 2009
91
92
H. Kondoh
12-dimethyl-benzanthracene (DMBA), followed by treatment with phorbol esters
(e.g., TPA), can result in skin tumors. In these experimental models, each step is
called initiation, promotion, and progression, respectively, supporting the notion that
cancer progression includes multistep events.
From the pathologic and clinical viewpoint, the process of tumorigenesis in
vivo can be divided into three or more stages: immortalization, transformation,
and metastasis [2]. Among other factors, the acquisition of unrestricted proliferative potential, called cellular immortalization, would be involved in the very early
stage in vivo. During immortalization, several biological events are supposed to be
required; bypassing senescence, evasion of apoptosis and antigrowth signals, growth
factor independence, and so on. Then, to grow more aggressively in vivo, these
immortalized cells should be transformed to be anchorage-independent, resistant to
contact inhibition, proangiogenic, and so forth. In the end stage, they easily detach
from each other, more easily attach to and degrade matrix components, and invade
and migrate to other tissues (metastasis).
All these properties are also distinct hallmarks of cancerous cells compared with
their normal counterpart [3]. Another candidate for inclusion in this list would be
enhanced glycolysis, noted by Otto Warburg more than seven decades ago. He first
reported that cancerous tissues or cells display increased glycolysis by an unknown
mechanism. A high glycolytic rate, even under high oxygen conditions, is referred to
as the Warburg effect. This property is well used in clinical practice for the detection
of metastatic tumor mass by positron-emission scanning of [18 F]fluoro-2-deoxy-Dglucose. Thus there is no dispute about its clinical significance, whereas there has
been a controversy as to exactly when cancer cells become highly glycolytic in
vivo. In other words, one big emerging question is the following: In which stage of
multistep oncogenesis would the enhancement of glycolysis be involved?
1.2 Enhanced Glycolysis Is Required During Transformation
to Adapt to Hypoxic Conditions
It was widely assumed that cancer cells maintain upregulated glycolytic metabolism
to adapt to the hypoxic conditions in vivo, as solid aggressive tumors outgrow the
blood supply of the feeding vasculatures [4]. Alternatively, it has been proposed
that the increased glucose flux might improve efficiency of glucose utilization in
a microenvironment in which glucose is limited. In such a context, the glycolytic
response represents a successful metabolic adaptation of cancer cells in vivo.
The concomitant induction of angiogenesis and enhancement of glycolysis with
cell proliferation is mediated partly by activating the hypoxia-inducible transcriptional factor (HIF-1) [5]. Hypoxia increases HIF-1␣ levels in most cell types, and
HIF-1 mediates adaptive responses to changes in tissue oxygenation. Thus, HIF-1
can directly upregulate expression of a set of genes involved in both local and global
responses to hypoxia, including angiogenesis, erythropoiesis, respiration, and most
of the glycolytic enzymes: Hexokinase 1 (HK1), Hexokinase 2 (HK2), Autocrine
The Warburg Effect and Immortalization
93
Motility Factor/Glucose-6-Phosphate Isomerase (AMF/GPI), Enolase 1 (ENO1),
glucose transporter 1 (GLUT1), glyceraldehyde-3-P dehydrogenase (GAPDH), lactate dehydrogenase (LDH), 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase
3 (PFKBF3), phosphofructokinase liver form (PFKL), phosphoglycerate kinase 1
(PGK1), pyruvate kinase muscle form (PKM), triose phosphate isomerase (TPI).
Interestingly, both HIF-1 and glycolytic enzymes are overexpressed in many tumors
and cancer cells. Moreover, ectopic expression of glycolytic enzyme LDH can transform culture cells, whereas the inactivation of LDH ablates the transformation ability in cancer cells [6]. Altogether, these data support a functional link between
enhanced glycolysis and cellular adaptation during tumor formation and expansion.
In conclusion, the transformation during multistep oncogenesis would require the
enhancement of glycolysis (Fig. 1).
1.3 Oncogene and Glycolysis: ras and c-Myc
However, the Warburg effect cannot be explained solely by cellular adaptation to
hypoxic conditions for the following reasons.
1. The cellular adaptation model does not explain the constitutive metabolic change
that maintains high glycolytic rates in cultured cancer cells even under 20% oxygen in vitro [7].
A
B
c-Myc
p53 KO
Immortalization
Immortalization
Glycolysis
Hypoxia
Transformation
HIF-1
Hypoxia
Transformation
Glycolysis
Metastasis
HIF-1
Glycolysis
Metastasis
Fig. 1 Two models of glycolysis involvement during multistep tumorigenesis.(A) Enhanced glycolysis during transformation. In the growing solid tumors, the core of tumors would suffer hypoxia
as they outgrow the feeding capacity of the neovasculature. To adapt to hypoxic conditions, HIF-1
is activated; which in turn enhances glycolysis. (B) Enhanced glycolysis during immortalization
and transformation. Other than the model in (A), glycolysis would be involved also in the earlier
step of immortalization. Immortalizing genes, induction of c-Myc, or knock-down of p53 (p53KO)
would enhance glycolysis in cancerous cells
94
H. Kondoh
2. Ectopic expression of HIF-1 in culture cells induces cell cycle arrest, which
argues against the proposition that glycolysis in cancer cells is regulated only
by HIF-1 or hypoxia.
3. Indeed, emerging evidence indicates that glycolysis is regulated in a cellular context–dependent manner and by multiple genetic factors: oncogenes (ras,
c-Myc), tumor suppressor genes (p53), signaling kinases (AMP kinase, Akt
kinase, Pak1 kinase), and so forth.
These facts suggest that there would be possible other mechanisms to increase
glycolysis during multistep oncogenesis other than adaptation to hypoxia. Interestingly, several groups reported that c-Myc plays a critical role in cellular immortalization of human primary epithelial cells and fibroblasts via telomerase activation
[8], and c-Myc could also enhance glycolysis through transcriptional upregulation
of several glycolytic enzymes: HK, PFK, TPI, GAPDH, ENO, and LDH [9]. Taken
together, it is possible that glycolysis could be enhanced in the earlier step of tumorigenesis before cellular transformation occurs.
2 Enhanced Glycolysis Is Also Required in Immortalization
2.1 Senescence Is a Barrier for Tumorigenesis
Historically, senescence and aging research came into their own around the 1960 s.
Hayflick and Moorhead reported replicative senescence in tissue cultured primary
cells in 1967 [10], and Harman proposed a radical theory of aging in 1956 [11].
Recently, there has been a resurgence in senescence biology, as cancer cells are
known to be immortal both in vivo and in vitro.
Most somatic cells have a limited replicative capacity under standard tissue culture conditions and suffer a permanent cell cycle arrest, called replicative senescence
[12]. Replicative senescence is induced by telomere erosion upon reaching replicative exhaustion, which can be bypassed by the ectopic expression of telomerase
in human fibroblasts. It is well established that senescence can also be induced in
a telomere-independent manner, called stress-induced senescence (SIS) [13]. Cells
suffering either replicative or stress-induced senescence (SIS) are phenotypically
similar: they adopt an enlarged and flattened appearance; deposit increased amounts
of extracellular matrix; express elevated levels of inhibitor of plasminogen activator
type 1 (PAI-1); develop lipofuscin granules, single prominent nuclei, senescenceassociated heterochromatin, senescence-associated ␤-galactosidase activity; and
show negligible DNA synthesis (Fig. 2).
Interestingly, although mouse telomeres are much longer than their human counterparts (60 kb vs. 12 kb), mouse embryonic fibroblasts (MEFs) stop growing
within a shorter time when compared with human primary fibroblasts [14]. Apart
from telomere structure, there are several other factors influencing cellular life
span, as some cell types reach a proliferative barrier long before telomere erosion
becomes critical. It is now recognized that cells can arrest virtually at any age, in
The Warburg Effect and Immortalization
95
Premature
senescence
Oxidative, or
Culture stress
glycolysis
declined
Replicative
senescence
Telomeric erosion
Immortal
Primary cells
Radical
Scavanger
reduced
ROS
Enhanced
glycolysis
Fig. 2 Glycolysis and cellular senescence. Senescence can be induced in telomere-dependent or
-independent manner. In both cases, glycolysis declines. Enhanced glycolysis can immortalize
primary cells with significant reduction of oxidative damage
a telomere-independent manner. For example, oncogenic stimuli can induce premature senescence, called oncogene-induced senescence. Thus, cellular senescence
also constitutes a potent anticancer mechanism [15].
2.2 Oxidative Stress and Senescence
Recent studies suggest that oxidative damage and the accumulation of reactive oxygen species (ROS) are closely involved in senescence. ROS, such as superoxide
anion and hydroxyl radical, are produced during cellular metabolism mainly in the
mitochondria. ROS are also produced in response to different environmental stimuli such as UV, IR, chemicals, hyperoxia, or hydrogen peroxide treatment. Abnormal ROS accumulation and its effects on intracellular macromolecules (oxidation of
lipid, protein, and DNA) induces cumulative damage at the cell, tissue, and organism level. Mild oxidative stress (e.g., treatment with low concentrations of hydrogen
peroxide) is enough to induce senescence in primary cells. Interestingly, premature
senescence induced by culture stress or oncogene-induced stress is associated with
oxidative damage in cells [16].
Increased ROS accumulation is also observed during replicative senescence. The
replicative potential of both murine and human fibroblasts is significantly enhanced
under low oxygen, associated with less oxidative damage inflicted than under
96
H. Kondoh
normoxia (O2 20%) [17]. Immortalized cells suffer less oxidative damage than do
primary fibroblasts when cultured under 20% O2 . Moreover, immortalized cells are
more resistant to the deleterious effects of hydrogen peroxide than are primary cells.
Thus, ability to resist oxidative stress could be a clue for explaining the immortality
of cancer cells.
Several radical scavengers can protect cells against oxidative stress. The superoxide
dismutase enzyme (SOD) converts superoxide anions into hydrogen peroxide, and
hydrogen peroxide can be detoxified by catalase. Consequently, these antioxidant
enzymes can impact the proliferation of both primary and immortal cells as they
can counteract ROS effects. The ability of SOD to bypass senescence has been well
studied and established in various cells or organisms. Increased expression of SOD
can extend the life span of primary fibroblasts [18]. Conversely, knockdown of SOD
using small interfering RNA (siRNA) induces premature senescence accompanied by
p53 activation. Transgenic fly overexpressing SOD [19] or the detoxifying enzyme
catalase [20] presents an extended organism life span. Although it is clearly established
that these antioxidant scavengers are essential for proliferation of immortal cells, little
is known to date on the specific regulation operating in cancer cells.
2.3 Senescence-Bypassing Effect of Enhanced Glycolysis
We recently found that glycolytic enzymes can modulate the cellular life span of
MEFs [21]. In a senescence-bypassing screening in MEFs using a retroviral cDNA
library, we isolated the glycolytic enzyme phosphoglycerate mutase (PGM). PGM
converts 3-phosphoglycerate to 2-phosphoglycerate during glycolysis. Analysis of
the impact of other glycolytic enzymes on senescence in MEFs showed that glucophosphate isomerase (GPI) could also drive immortalization of MEFs. Ectopic
expression of PGM or GPI increases glycolytic flux, decreases the oxidative damage
that MEFs are exposed to, and extends the life span of primary MEFs. Conversely,
knockdown of PGM or GPI via specific siRNA induces premature senescence.
Moreover, we and others found that the glycolytic flux declines during senescence
both in murine and human fibroblasts [22].
How can an increase in glycolysis immortalize primary cells? From the data
presented above, it appears that enhanced glycolysis can protect cells from oxidative
stress and as a consequence avoid senescence [21]. MEFs immortalized by PGM
or GPI suffer less oxidative damage than do control cells as estimated by cytosolic
ROS staining, or quantification of 8-hydroxydeoxyguanosine (8-OHdG), a hallmark
of oxidative DNA damage lesions.
Moreover, mouse embryonic stem (ES) cells also present a surprisingly high glycolytic rate [23]. Thus similarly to cancer cells, ES cells display the Warburg effect.
ES cells are an exception that can bypass culture stress–induced senescence among
other primary cells. We hypothesized that reversible modifications such as a specific
factor playing a role in ES cell immortality enhances the glycolytic rate resulting
in an increased protection from oxidative stress. This metabolic protection might
contribute to preserve the genome integrity of ES cells thereby avoiding genetic
The Warburg Effect and Immortalization
97
alterations and allowing them to maintain their self-renewal capacity. Also, these
metabolic rates can be reversed, which would explain why differentiated cells do
not present this enhanced metabolic level.
2.4 Glycolysis as Radical Scavenger
How can enhanced glycolysis protect cells from oxidative damage? One clue is the
common metabolic feature shared between several immortalized cells: cancer cells,
immortalized primary cells, and ES cells [22]. All these cells display enhanced glycolysis with decreased mitochondrial respiration. Thus, increased glycolysis during immortalization could be accompanied by the downregulation of mitochondrial
function by unknown mechanism(s) and would result in less intrinsic ROS production, which will be discussed later.
A second possible mechanism is that increased glycolysis could imitate or activate radical scavengers. Interestingly, the antioxidant function of some scavengers
(such as reduced gluthation (GSH) or Thioredoxin (TRX)) is closely coupled to
the NADPH/NADP balance. Most of the NADPH/NADP is produced through the
pentose phosphate pathway (PPP), a branching metabolic pathway derived from the
glycolytic pathway. Enhanced glycolysis might activate PPP and increase the level
of NADPH as by-products (Fig. 2).
Although this remains a hypothesis, several reports support it. The impact of
glucose-6-phosphate dehydrogenase (G6PD) activity on cell proliferation is well
established [24]. G6PD catalyzes the rate-limiting step in the pentose phosphate PPP,
which is responsible for the recycling of NADPH and maintenance of the redox balance as described above. G6PD-deficient human fibroblasts have a reduced life span
that is attributable to oxidative stress and can be corrected by the ectopic expression
of this enzyme [25]. Both G6PD activity and the NADPH pool decline during continued culture passage, presumably as a consequence of the accumulation of oxidative
damage. Importantly, ES cells ablated from G6PD expression are extremely sensitive
to oxidative damage, showing massive apoptosis at low concentration of oxidants that
are nonlethal for wild-type ES cells. It would, therefore, be worth exploring in the
future whether enhanced glycolysis can then promote increased NADPH production
via the PPP thereby exerting its antisenescence function.
3 Opposing Effect of Glycolysis on Life Span Observed
in Different Organisms
3.1 Discovery of Sir2 as Longevity Gene in Yeast
Any discussion about longevity and glycolysis merits evaluation of the results
observed in yeast. The hypothesis that enhanced glycolysis can promote proliferation would not be applied in every species. It seems quite opposite in the case
98
H. Kondoh
of budding yeast. In yeast, Sir2 gene was isolated as a longevity gene, which can
be activated in a calorie-restricted condition [26]. Calorie restriction is also wellknown to extend the life span of mouse and yeast. In a calorie-restricted condition,
glycolysis declines in yeast, accompanied by an elevation of the respiratory rate.
Mutations in the glycolytic pathway have also been reported to extend life span in
yeast, suggesting that decreased glycolysis has a causal effect on extension of yeast
life span.
Sir2 works as an NAD-dependent histone deacetylase. In yeast, restricting food
availability affects two pathways that activate Sir2 enzymatic activity [27]. First,
calorie restriction turns on a gene called PNC1, which produces an enzyme that rids
cells of nicotinamide, a small molecule similar to vitamin B3 that normally represses
Sir2. Second, activated respiration during calorie restriction is a mode of energy
production that creates NAD as a by-product and lowers NADH. In consequence,
NAD-dependent Sir2 activity is induced by calorie restriction. Thus, Sir2 is a key
molecule linking calorie restriction and longevity in yeast.
3.2 Calorie Restriction and Sir2
It seems that glycolysis has an opposing effect on yeast life span, in sharp contrast
with the Warburg effect observed in cancerous or immortalized mammalian cells. To
date, it is quite difficult to understand these differences in glycolytic profile between
long-lived yeast in calorie restriction and cancer cells in tissue culture, but there are
some possible explanations.
First, the mammalian homolog of Sir2 gene, called Sirt1, has less impact on
aging of mammalian primary cells in tissue culture. Sirt1 can directly bind and modify several DNA binding proteins: histone, p53, Peroxisome Proliferator-Activated
Receptor-␥ Coactivator-1 (PGC-1), and possibly other unknown transcriptional
factors. At least p53 is not conserved in yeast. Mammalian Sirt1 might target
more unknown different factors, which would mask its senescence-bypassing effect
observed in yeast.
Second, aging of budding yeast is quite different from senescence observed in
primary culture cells. The mother yeast cells suffer the accumulation of budding
scars on their surface during the division process. In parallel, aging yeast cells
show ballooning nucleolus, which is caused by accumulation of circular ribosomal DNAs [28]. These aging phenotypes are quite specific to yeast, which might
explain the difference of aging mechanisms between yeast and mammalian culture
cells.
Third, Sir2 could affect mammalian organismal aging rather than senescence of
mammalian culture cells. Resveratrol is a chemical reagent that activates Sirt1 and
is enriched in red wine. Resveratrol improves health and survival of mice on a highcalorie diet. In this sense, calorie restriction–induced longevity observed in yeast
should be more relevant to mammalian organismal longevity in calorie-restricted
condition rather than the Warburg effect observed in cancer cells.
The Warburg Effect and Immortalization
A
99
B
Calorie
restriction
Glycolysis declined
Calorie
restriction
Sir2 activated
Others?
Respiration increased
Sir2 activated
Longevity
Glycolysis?
Longevity
Fig. 3 Impact of calorie restriction and Sir2 on yeast longevity and mammalian organismal aging.
(A) In budding yeast, life span is extended in calorie-restricted condition. In such a case, glycolysis
declined associated with enhanced respiration. Enhanced respiration will promote NAD production, leading to activation of Sir2. (B) In mouse model, calorie restriction can delay organismal
aging. In contrast, high-calorie diet can shorten the life span, which would be restored by treatment with resveratrol, activating chemicals for Sir2. The significance of glycolysis in this situation
is still not clear
Collectively, Sir2 is a highly conserved protein and serves as a key to modulate
metabolic switching. Sir2 would work as a longevity gene in calorie-limited conditions, but its effect on replicative and stress-induced senescence remains to be
clarified (Fig. 3). Thus, yeast might not be a good model to explore and explain
the Warburg effect that occurs primarily in cancerous and immortalized mammalian
cells.
4 Novel Regulatory Mechanism of Glycolysis
4.1 p53 and Glycolysis
Recent advances in understanding the regulatory mechanism of glycolysis also supports our hypothesis that the Warburg effect is a very early event in tumorigenesis.
Inactivation of tumor suppressor gene p53 increases glycolytic flux in vitro andin
vivo. Ablation of p53 significantly downregulates mitochondrial respiration in vitro
and in vivo.
p53 is the most frequently mutated gene in various types of cancer and functions
as a transcriptional factor to induce cell cycle arrest, apoptosis, and so forth. Inactivation of p53 immortalizes primary cells in vitro, implicating that p53 can also
affect the senescence process. p53 knockout mice are cancer prone, possibly due to
this senescence effect. Partial activation of p53 induces premature organismal aging
100
H. Kondoh
Extra single copy of p53
Longevity with less tumor events
Activated allele of p53
Premature ageing
p53 KO
Short lifespan due to cancer prone
Old wild-type
Cancer prone
Fig. 4 p53 and organismal life span. p53 null mice are cancer prone, suffering short life span. p53
activating allele knock-in mice display premature aging phenotype. In contrast, mice knocked-in
for an extra copy of p53 acquire organismal longevity due to less oxidative damage and fewer
tumor events KO-knock out
in vivo[29] (Fig. 4). One possible interpretation would be that one of the major targets of p53 involved in senescence induction impacts metabolic regulation, which
renders cells sensitive to oxidative stress. TP53-induced glycolysis and apoptosis
regulator (TIGAR) may be such a target through which p53 can modulate oxidative
stress in vivo. Recent works suggest that mice knocked-in for an extra copy of p53
acquire longevity associated with a resistance to oxidative damage [30]. It is possible that the metabolic shift caused by p53 might modulate cellular senescence and
organismal aging through reduction of oxidative damage in vivo and in vitro.
p53 regulates both glycolysis and mitochondrial respiration through different targets. p53 can target glycolytic enzymes (HK and PGM) and modulate the overall
flux of glycolysis in primary cells. On the other hand, p53 knockdown can decrease
mitochondrial respiration [23]. This is mainly due to inhibition of cytochrome c
oxidase 2 (SCO2) [31], a direct target of p53. SCO2 is critical in regulating the
cytochrome c oxidase complex, the major site of oxygen utilization. In this way, p53
could modulate both glycolysis and respiration in a concerted manner and would
affect cellular life span. This fact suggests that stable alterations at the genetic or
epigenetic levels, including inactivation of p53, provide a possible explanation for
the enhanced glycolysis that cancer cells present in vitro.
4.2 PGM and Its Regulation
Since Dr. Warburg’s discovery, research on glycolysis has been mainly performed by
using established cancer and muscle cell lines. But these cells are already immortal
with several genetic alterations and presumably not suitable to dissect the impact of
immortalizing genes, such as c-Myc or p53 knockdown, on the glycolytic pathway.
For example, phosphofructokinase (PFK) is described as the rate-limiting
enzyme for glycolysis, while its overexpression has no effect on glycolytic flux
The Warburg Effect and Immortalization
101
in immortalized and nonimmortalized cells. Ectopic expression of other enzymes,
PGM or GPI, can increase glycolytic flux drastically in primary cells, indicating that
regulation of glycolysis pathway is largely cellular context–dependent. For example,
PGM protein levels are tightly regulated by the ubiquitination pathway in primary
cells, while they are not so regulated in immortalized cells. Chemical screening of
a breast cancer drug identified the PGM inhibitor as the most potent one, suggesting again that PGM can be a key step in the glycolytic pathway in a tissue-specific
or cellular context–dependent manner. Interestingly, HIF-1 is a key transcriptional
factor that upregulates many glycolytic enzyme genes under hypoxic conditions,
but PGM is the only exception. The regulatory mechanism for PGM remains to be
clarified.
5 Conclusion
Glycolysis serves not only as an energy source for cells but also is essential in the
steps of immortalization and transformation during multistep tumorigenesis, as its
enhancement can render cells resistant both to oxidative stress and hypoxic conditions, respectively. Cancerous cells show a concerted metabolic shift, including
enhanced glycolysis with reduced mitochondrial respiration by an as yet poorly
characterized mechanism. The regulation of glycolysis relevant to the senescence
process is probably more complicated than we have expected. Future work to clarify its molecular mechanism should be employed in various cell types, including
primary cells and ES cells—not just in cancer cells. p53 might be a key molecule as
it plays a significant role both in organismal and cellular aging. These findings will
also contribute to improve and identify new anticancer therapies in the future.
References
1. Wu X, Pandolfi PP. Mouse models for multistep tumorigenesis. Trends Cell Biol 2001; 11:
S2–S9.
2. Braithwaite KL, Rabbitts PH. Multi-step evolution of lung cancer. Semin Cancer Biol 1999;
9:255–265.
3. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000; 100:57–70.
4. Dang CV, Semenza GL. Oncogenic alterations of metabolism. Trends Biochem Sci 1999;
24:68–72.
5. Semenza GL. Hypoxia-inducible factor 1: master regulator of O2 homeostasis. Curr Opin
Genet Dev 1998; 8:588–594.
6. Shim H, Dolde C, Lewis BC, et al. c-Myc transactivation of LDH-A: implications for tumor
metabolism and growth. Proc Nat Acad Sci USA 1997; 94:6658–6663.
7. Gatenby RA, Gillies RJ. Why do cancers have high aerobic glycolysis? Nat Rev Cancer 2004;
4:891–899.
8. Wang J, Xie LY, Allan S, Beach D, Hannon GJ. Myc activates telomerase. Genes Dev 1998;
12:1769–1774.
9. Kim JW, Zeller KI, Wang Y, et al. Evaluation of myc E-box phylogenetic footprints in glycolytic genes by chromatin immunoprecipitation assays. Mol Cell Biol 2004; 24:5923–5936.
102
H. Kondoh
10. Hayflick L, Moorhead PS. The serial cultivation of human diploid cell strains. Exp Cell Res
1961; 25:585–621.
11. Harman D. Aging: a theory based on free radical and radiation chemistry. J Gerontol 1956;
11:298–300.
12. Wright WE, Shay JW. Historical claims and current interpretations of replicative aging. Nat
Biotechnol 2002; 20:682–688.
13. Sherr CJ, DePinho RA. Cellular senescence: mitotic clock or culture shock? Cell 2000;
102:407–410.
14. Serrano M, Blasco MA. Putting the stress on senescence. Curr Opin Cell Biol 2001; 13:
748–753.
15. Campisi J. Cellular senescence as a tumor-suppressor mechanism. Trends Cell Biol 2001;
11:S27–S31.
16. Serrano M, Lin AW, McCurrach ME, Beach D, Lowe SW. Oncogenic ras provokes premature
cell senescence associated with accumulation of p53 and p16INK4a. Cell 1997; 88:593–602.
17. Itahana K, Zou Y, Itahana Y, et al. Control of the replicative life span of human fibroblasts by
p16 and the polycomb protein Bmi-1. Mol Cell Biol 2003; 23:389–401.
18. Serra V, von Zglinicki T, Lorenz M, Saretzki G. Extracellular superoxide dismutase is a major
antioxidant in human fibroblasts and slows telomere shortening. J Biol Chem 2003; 278:
6824–6830.
19. Parkes TL, Elia AJ, Dickinson D, Hilliker AJ, Phillips JP, Boulianne GL. Extension of
Drosophila lifespan by overexpression of human SOD1 in motorneurons. Nat Genet 1998;
19:171–174.
20. Orr WC, Sohal RS. Extension of life-span by overexpression of superoxide dismutase and
catalase in Drosophila melanogaster. Science 1994; 263:1128–1130.
21. Kondoh H, Lleonart ME, Gil J, et al. Glycolytic enzymes can modulate cellular life span.
Cancer Res 2005; 65:177–185.
22. Kondoh H, Lleonart ME, Gil J, Beach D, Peters G. Glycolysis and cellular immortalization.
Drug Discov Today 2005; 2:263–267.
23. Kondoh H, Lleonart ME, Nakashima Y, et al. A high glycolytic flux supports the proliferative
potential of murine embryonic stem cells. Antioxid Redox Signal 2007; 9:293–299.
24. Tian WN, Braunstein LD, Apse K, et al. Importance of glucose-6-phosphate dehydrogenase
activity in cell death. Am J Physiol 1999; 276:C1121–C1131.
25. Ho HY, Cheng ML, Lu FJ, et al. Enhanced oxidative stress and accelerated cellular senescence
in glucose-6-phosphate dehydrogenase (G6PD)-deficient human fibroblasts. Free Radic Biol
Med 2000; 29:156–169.
26. Sinclair DA, Lin SJ, Guarente L. Life-span extension in yeast. Science 2006; 312:195–197;
author reply 195–197.
27. Sinclair DA, Guarente L. Unlocking the secrets of longevity genes. Sci Am 2006; 294:48–51,
54–57.
28. Sinclair DA, Guarente L. Extrachromosomal rDNA circles – a cause of aging in yeast. Cell
1997; 91:1033–1042.
29. Tyner SD, Venkatachalam S, Choi J, et al. p53 mutant mice that display early ageingassociated phenotypes. Nature 2002; 415:45–53.
30. Matheu A, Maraver A, Klatt P, et al. Delayed ageing through damage protection by the
Arf/p53 pathway. Nature 2007; 448:375–379.
31. Matoba S, Kang JG, Patino WD, et al. p53 regulates mitochondrial respiration. Science 2006;
312:1650–1653.
Metabolic Modulation of Carcinogenesis
Shireesh P. Apte and Rangaprasad Sarangarajan
Abstract Emerging evidence inextricably links cellular pathways that involve
glucose metabolism (via oxidative phosphorylation and glycolysis), oxygen sensing, apoptosis, cell cycle regulation, and immune recognition and response. This
accounts for the epidemiologic and epigenetic effects of calorie intake and quality, obesity, diabetes, immunosuppression, chronic infection or inflammation, aging,
and viral diseases on carcinogenic rates and trends. The mitochondrion is a critical
organelle that coordinates and integrates the above-mentioned signaling pathways
so as to ultimately affect cell proliferation or cell death. The differential regulation of these pathways not only depends on the varying levels of stress induced
by the above-mentioned epigenetic factors but also on acquired genetic mutations,
thereby attempting the prediction of cell survival or death, an exceedingly difficult proposition. Concepts that have not received much attention in the literature—
such as the role of the exogenous cytochrome C/NADH pathway in apoptosis and
oxidative phosphorylation; the switch to glycolysis at lower pyruvate thresholds in
cancer cells and the paradox that this glycolytic switch may be necessary to provide ATP and matrix alkalinization necessary for apoptosis to occur; reconciling the
observations that caloric restriction causes a generalized decrease in apoptosis yet
incidences of cancer are significantly decreased in calorie-restricted animals; the
unsaturation patterns of the fatty acids comprising the mitochondria-specific phospholipids; and the variation of the intermediary metabolism of tumors—with time
will be addressed from the viewpoint of altered cellular respiration. The implications
of attenuation of cellular respiration on the mitochondrial adhesion and localization
of key immunoregulatory and glycolytic-apoptotic enzymatic or protein complexes
will be discussed.
Keywords Glycolysis · Oxidative phosphorylation · Apoptosis · Carcinogenesis ·
Cell cycle · Hypoxia · Warburg · Calorie restriction · Cytochrome c
S.P. Apte (B)
Alcon Laboratories, 6201 South Freeway, Forth Worth, Texas 76134, USA
e-mail: Shireesh.apte@alconlabs.com
S.P. Apte, R. Sarangarajan (eds.), Cellular Respiration and Carcinogenesis,
DOI 10.1007/978-1-59745-435-3 8,
C Humana Press, a part of Springer Science+Business Media, LLC 2009
103
104
S.P. Apte, R. Sarangarajan
Oxidative phosphorylation (OXPHOS), the last of the major steps in aerobic respiration and energy generation, occurs in the mitochondrion, an intracellular organelle
that also controls and modulates apoptosis. A dysfunctional OXPHOS or an inadequate oxygen supply can therefore modulate apoptotic function. It can also increase
the relative or absolute contribution of an oxygen-independent pathway of respiration and energy generation—glycolysis—to meet cellular energy requirements.
This adaptive response temporally progresses from being proapoptotic to becoming antiapoptotic in nature. The proapoptotic effect may predominate in the earlier
time period/stage of adaptation if tumor suppressors and apoptotic machinery are
functional. The antiapoptotic effect may predominate in the latter times/stages of
adaptation if either the tumor suppressors and apoptotic machinery is dysfunctional
or is made so by this adaptive response itself.
1 A Decreased OXPHOS Phenotype Attenuates Apoptosis
A decreased content of OXPHOS complexes II, III, and IV, reduced level of ATPsynthase, and suppression of NDUFA1, an accessory component of OXPHOS, are
all associated with the development of tumors, increased tumor aggressiveness,
and antiapoptotic effects. Deficiencies in mtETC, ATP synthesis machinery (F0 F1 ATPase), or ANT abrogate the apoptotic effects of Bax. OXPHOS dysfunction is
associated with a proinvasive and pro-proliferative modeling of the ECM by altering transcriptional regulation of genes coding for members of MMP/TIMP, UPA,
and CCN in human osteosarcoma cells.
An impaired mtETC may increase cellular NADH levels because electrons can
no longer be transferred as efficiently to oxygen and NADH cannot be oxidized
back to NAD+ . Increased levels of NADH induce protection from apoptosis and
decreased cell adhesion by promoting the binding of the tumor suppressor E1A to
the (NADH activated) corepressor CtBP.
Progression toward a more tumorigenic state is associated with increased and
decreased sensitivity to 2-deoxyglucose and oligomycin, respectively, consistent
with an increase and decrease in glycolysis and OXPHOS, respectively [1].
Dysfunction or mutations in tumor suppressors may cause attenuation in
OXPHOS. A decrease in COX activity due to p53 attenuation can lead to a diminished OXPHOS phenotype [2]. In hepatomas caused by mutated p53, the mitochondrial porin bound hexokinase II is increased [3]. This enzyme isoform binds to the
mitochondrial membrane through interaction with VDAC and inhibits apoptosis by
suppressing the release of intramembrane space proteins in part by preventing the
opening of the PTP [4]. In CCRC, VHL deficiency is one of the factors responsible
for downregulation of the biogenesis of OXPHOS complexes, as well as for upregulation of HIF-1␣. AIF deficiency has been shown to produce a reduced OXPHOS
phenotype.
Metabolic Modulation of Carcinogenesis
105
2 Reactive Oxygen Species and Carcinogenesis
A general picture that emerges from the literature is that cellular ROS can modulate
cell-signaling pathways, carcinogenesis, or cell death depending upon their cellular
levels. Whether a particular cellular level of ROS triggers proliferation or apoptosis
also seems to depend on the magnitude of the cellular stress—which in turn depends
on substrate and oxygen supply, the existence of mutations in key tumor suppressor
genes, OXPHOS function, and the local cellular environment including the levels of
antioxidant enzymes such as SOD and catalase.
Mitochondria can generate ROS not only accidentally but also via a regulated
enzymatic mechanism, thereby attesting to the importance of ROS in apoptosis. This
mechanism involves the electron transfer between the proapoptotic signal p66Shc
and cyt-C that generates ROS, which in turn triggers apoptosis by opening the PTP
and subsequent caspase activation.
Interestingly, the redox reaction between p66Shc and cyt-C is favored by an excess
of reduced cyt-C; as may occur under hypoxia when COX activity is decreased.
Proapoptotic molecules such as Bax and tBid cause an increase of free intermembrane cyt-C, which is then fully reduced by the rotenone-insensitive NADH-cyt b5
reductase (see Section 8). Therefore, apoptotic induction in attenuated OXPHOS
or hypoxic cells routes through the exogenous NADH/cyt-C electron transport
pathway.
Mitogenic signals during low-stress conditions activate E2F—which in turn
decreases ROS—so as to induce proliferation. In conditions of high cellular stress,
however, tumor suppressor genes such as p53 and p16INK4a are activated, leading to
inhibition of E2F. Consequently, ROS levels accumulate to senescence-promoting
levels. High levels of ROS also induce a PKC-mediated block of cytokinesis.
Under conditions where either OXPHOS or ROS scavenging is dysfunctional,
increased ROS levels may inactivate certain tumor suppressors such as PTEN leading to increased activation of Akt and its downstream targets, HIF-1␣ and VEGF.
The activation of the Akt oncogene has been demonstrated to be sufficient to
stimulate the switch from OXPHOS to aerobic glycolysis, in part by increasing
mitochondria-associated hexokinase activity. Akt is a negative regulator of AMPK
activity and hence can overcome the tumor-suppressor activity of LKB1. ROS mediated impaired OXPHOS induced PTEN oxidation may also suppress p53 activity by
allowing nuclear translocation of Mdm2.
3 The Hexosamine Biosynthetic Pathway
Adaptive glycolysis triggered by impaired OXPHOS, increased ROS generation,
hypoxia, or deregulation of nutrient uptake pathways may siphon more glucose into
the hexosamine biosynthetic pathway (HBP). Apart from causing insulin resistance,
an increased flux through the HBP may cause increased O-GlcNAc glycosylation
106
S.P. Apte, R. Sarangarajan
and inactivation of the proapoptotic protein Bnip3 [5], thereby conferring a survival
advantage on glycolytic cells [6].
Furthermore, increased posttranslational modification of transcriptional factors
such as Sp1—mediated by O-GlcNAc glycosylation—may result in downregulation
of OXPHOS gene expression and decreased mitochondrial function by impairing
the activity or expression of known regulators of OXPHOS transcription PGC-1␣
and NRF [7]. This phenomenon may further exacerbate the shift from OXPHOS to
glycolysis.
4 Diversion of Pyruvate from Mitochondria
The prevalence of the Crabtree effect (a preference to ferment glucose into ethanol)
in the bakers’ yeast, Saccharomyces cerevisiae, has been attributed in part to differences in enzymes at the branch point between respiration (PDH) and fermentation
(PDC). The concentration of pyruvate that gives half maximal activity (Km ) of PDH
is lower than that for PDC. The latter enzyme exhibits cooperativity with substrate concentration and also has a higher maximal activity. The net result is to
make metabolic flux through PDH favored at low pyruvate concentrations and flux
through PDC favored at higher ones. It is instructive to compare aerobic metabolism
in this organism to that exhibited by tumor cells.
A decreased NAD:NADH ratio, which occurs in part as a consequence of
impaired mtETC and which is characteristic of highly glycolytic malignant cells,
downregulates the PDH complex [8] so that entry of pyruvate into the TCA cycle is
impaired.
The unusual metabolite, acetoin, unique to tumors and formed by the nonoxidative decarboxylation of pyruvate with acetic acid, is a competitive inhibitor of
intramitochondrial tumor PDH. Km values for pyruvate for the PDH of tumor cell
lines were significantly lower (up to threefold) than were those for the PDH of
non-transformed cells [9]. Broadly interpreted, the concentration of pyruvate below
which metabolic flux is favored through OXPHOS may be lower in tumor cells,
thereby contributing toward an oxygen supply–independent “pyruvate siphoning”
into glycolysis. This interpretation also implies that changes in metabolic enzyme
activity/polymorphism may be sufficient to initiate the “switch to glycolysis” independent of the cellular oxygen concentration.
Nonhypoxic activation of HIF-1 through loss of VHL or activation of Akt or
other signal transduction pathways could increase PDK1 and trigger the Warburg
effect. HIF-1–mediated increased PDK1 activity has been shown to inhibit PDH,
causing the shunting of pyruvate from the mitochondria and subsequently attenuating mitochondrial respiration and ROS production in MEFs.
The influx of respiratory substrate into mitochondria has been shown to be an
important determinant of cell sensitivity to oxidant-induced apoptosis. Reduced
OXPHOS may hence induce low sensitivity to oxidative stress.
Metabolic Modulation of Carcinogenesis
107
5 Organelle and Cell Membrane Perturbations
Complexation of the mitochondria-specific phospholipid cardiolipin (CL) with cytC enables the latter to act as a CL-oxygenase, which is activated during apoptosis
and selectively oxidizes CL. Oxidized CL is required for the release of proapoptotic factors from the mitochondria into the cytosol. Progression of the apoptotic
program is hence dependent on the susceptibility of CL to oxidation; which in turn
is dependent on its polyunsaturation pattern [10]. Consistent with this observation,
enrichment of CL with the highly unsaturated fatty acid docosahexaenoic acid conferred HL-60 cells with increased sensitivity to apoptosis induced by staurosporine
[11].
Hypoxia may be partly the effect of the incorporation of adulterated, nonoxygenating, or oxidized polyunsaturated fatty acids into the phospholipids of cellular
and mitochondrial membranes [12]. Such incorporation is hypothesized to cause
changes in membrane permeability that impair oxygen transmission into the cell.
Trans fats, partially oxidized PUFA entities, and inappropriate omega 6:omega 3
ratios are all potential sources of unsaturated fatty acids that can disrupt the normal membrane structure and oxygen diffusivity. Significant differences have been
observed between the phospholipid fatty acid profiles of the human colorectal
mucosa among the control, adenoma, and carcinoma groups [13].
Acetoin has been shown to have a potentiating effect on citrate formation, which
in turn is responsible for cytoplasmic cholesterol synthesis from the truncated
tumoral Krebs cycle [14]. Deregulation of cholesterol synthesis has been observed
in various cancer cell lines leading to its accumulation in every type of cell membrane. Cholesterol enrichment in membranes induces changes in biophysical properties, such as increased rigidification and decreased passive proton permeability. The
accumulation of cholesterol, in combination with altered ratios of omega 6:omega
3 fatty acids, partially oxidized PUFA entities, and trans-esters of fatty acids have
been hypothesized to impair oxygen transmission into the cell.
6 How Calorie Restriction Inhibits Carcinogenesis
Nowhere is the link between metabolism and carcinogenesis more profoundly
demonstrated than in studies of caloric restriction. CR has unequivocally been associated with increased life span and decreased carcinogenesis. In contrast, metabolic
disorders such as obesity [15] and diabetes [16] are known risk factors for the development of cancer. CR has been shown to promote mitochondrial biogenesis and
increase bioenergetic efficiency by modulating eNOS, PGC-1␣, and SIRT1 expression [17]. IGF1 inhibition by ER downregulates the longevity inhibitory gene product p66Shc and upregulates the induction of NRF1 and NRF2␤␥ mRNA.
The expression of mammalian SIRT1 is induced in CR rats as well as in
human cells that are treated with serum from such animals. SIRT1 deacetylates the
DNA repair factor ku70, causing it to sequester the proapoptotic factor Bax away
108
S.P. Apte, R. Sarangarajan
from mitochondria, thereby inhibiting stress-induced apoptotic cell death. CR rats
showed less apoptosis in skeletal muscle compared with their age-matched cohorts.
CR produces a reduction in body temperature of rodents and nonhuman primates. This has been suggested to occur by an increase in the coupling of oxidative
phosphorylation to ATP synthesis. Energy-deprived HepG2 hepatoma cells showed
increased mitochondrial biogenesis and OXPHOS. OXPHOS efficiency, as indicated by a higher ADP:O ratio, lower H2 O2 production, and lower mtDNA content
was found to correlate with Drosophila simulans survival [18]. Nutrient-induced
downregulation of OXPHOS genes in skeletal muscle has been suggested to represent a thrifty response to nutrient excess, which attempts to limit the compensatory
increase in energy expenditure normally designed to counterbalance the increase in
energy intake.
If caloric restriction causes a generalized decrease in apoptosis, then it should
be conducive to carcinogenesis. However, incidences of cancer are significantly
decreased in calorie-restricted animals. This apparent paradox can be explained by
invoking the hypothesis that growth advantage for mutated cells (along with the
probability of monoclonality and malignancy) is considerably increased (in part
due to decoupling and consequent deregulation of nutrient intake pathways independent of the extracellular matrix) if the normal multiclonal environment is damaged by a pathologic proapoptotic process that spares the apoptosis-resistant clones.
If the multiclonal environment is more stress resistant and robust—as would be
expected to be the case with CR—the selective growth advantage of mutated cells
is diminished.
As an example, senescence creates a local tissue environment that promotes the
growth of initiated or preneoplastic epithelial cells both in culture and in vivo [19].
To illustrate this effect further, CR results in a negative regulation of p53mediated transcriptional activation due to its deacetylation by SIRT1. A slight reduction in p53 activity permits the expression of several genes involved in cell-cycle
arrest (p21WAF1/CIP1 ), DNA repair (p53R2), and those involved in protection against
oxidative stress (TIGAR, sestrins, GPX1), thereby allowing cell survival. In contrast; a proapoptotic process mediated by elevated levels of p53 leads to the induction of pro-oxidant genes (PIG3, praline oxidase), the repression of transcription of
antioxidant genes (PGM, NQO1), and results in elevated intracellular ROS levels
and cell death or senescence in nonmutated cells or tumorigenesis in mutated or
apoptosis-resistant clones. This differential regulation of target genes in response to
varying levels of stress seems to be generally applicable also to ROS and Ca2+ in
the context of cell survival or death.
Apoptosis in the earlier stage of carcinogenesis may be necessary to increase
the selective pressure on the initiated cells to favor the emergence of proliferative,
apoptosis-resistant cells. For example, it has been hypothesized that Bax toxicity
may be a result of a Bax-induced impairment in the ability to perform OXPHOS.
Consistent with this hypothesis, it has been reported that Bax-induced lethality is
diminished under conditions that favor fermentation and that the cells that recover
from growth arrest after Bax expression frequently exhibit a permanent loss of respiration competence [20]. Deregulated activity of bcl-2 family members enhances
Metabolic Modulation of Carcinogenesis
109
tumor formation when overexpressed in the upper layers of the epidermis but retards
tumor formation when overexpression is restricted to the basal layer.
Furthermore, apoptosis-induced selection pressure in the early stage of carcinogenesis is supported by the finding that coexpression of two proapoptotic genes
(transforming growth factor ␤1 and the hepatitis B virus X) potentiates c-Myc–
induced hepatocarcinogenesis.
Clonal expansion of potentially malignant p53-deficient cell variants was suppressed in Bcl-2 overexpressing tumors, presumably due to the lack of selective
advantages for p53-deficient cells inside a population of cells resistant to apoptosis.
There is an age-related increase in the proportion of liver mitochondria with
fragile outer membranes as evidenced by an increased release of adenylate
kinase in hypotonic medium. Peroxidized cardiolipid in the mitochondrial membrane lowers the threshold of Ca2+ for PTP induction and cyt-C release. Aging
mice exhibit enhanced PTP activation in lymphocytes, brain, and liver. These
age-related, chronic, mitochondrial degenerative, apoptosis-favoring phenomena
further increase the probability of growth advantage of a subset of apoptosisresistant mutant clones.
That may be why pathologic proapoptotic processes such as paroxysmal nocturnal hemoglobinuria, myelodysplastic syndromes, chronic myeloid leukemia,
secondary acute leukemias, and immunosuppression-related non–Hodgkin’s lymphomas may be interpreted as “opportunistic” clonal and malignant diseases [21].
CR may cause a generalized increase in efficiency of OXPHOS because it causes
a decreased respiratory substrate flux through the HBP and decreased levels of UDPGlcNAc, the main substrate for protein glycosylation (see Section 3).
7 Chronic Infection and Carcinogenesis
Hypoxia is a characteristic feature of the tissue microenvironment during bacterial
infection. HIF-1␣ has been shown to be induced by bacterial infection, even under
normoxia, and regulates the production of key immune effector molecules, including granule proteases, antimicrobial peptides, nitric oxide, and TNF-␣. Mice lacking HIF-1␣ in their myeloid cell lineage showed decreased bactericidal activity and
failed to restrict systemic spread of infection from an initial tissue focus. Conversely,
activation of the HIF-1␣ pathway through deletion of VHL tumor-suppressor protein or pharmacologic inducers supported myeloid cell production of defense factors and improved bactericidal capacity. Therefore, any chronic infectious condition,
whether or not triggered by pathogens, may have the ability to upregulate HIF in the
vicinity of the affected cells.
Among agents that have been reported to cause uncoupling of OXPHOS to
ATP synthesis or a decrease in OXPHOS are Helicobacter pylori and hepatitis B
and the Epstein-Barr viruses; these are also selectogens for non–Hodgkin s lymphoma, liver cancer, and Burkitt s lymphoma, respectively. Epstein-Barr virus has
also been shown to induce synthesis of HIF-1␣, and the protein is also upregulated in
110
S.P. Apte, R. Sarangarajan
macrophages at normoxia after exposure to bacteria. Uncoupling protein 2 (UCP2)
expression is increased in most human colon cancers, and the level of expression
correlates with the degree of neoplastic changes. Bacterial lipopolysaccharide has
been shown to increase the expression of mitochondrial UCP2 in several tissues.
Differential compartmentalization of signaling molecules in cells is emerging as
an important mechanism for regulating the specificity of signal transduction pathways. Altered stress tolerance (and therefore altered heat shock protein levels) in
cells with dysfunctional OXPHOS can affect protein folding and thereby disrupt
their membrane localization or rate of ubiquitination [22]. Candidates subject to
such a pathologic response include the antiviral protein MAVS, whose disruption
of mitochondrial localization renders it unable to activate NF-␬B or IRF3 to induce
type-I interferons. The resultant attenuation of innate immunity can sensitize the
cell to carcinogenesis [23].
Increased MHC class I expression serves to alert the immune system to cells with
mitochondrial mutations [24]. Mice deficient in MHC I accumulate mitochondrial
DNA deletions in various tissues. Therefore, immunocompromised individuals may
not be effectively able to eliminate cells containing mitochondrial DNA mutations.
Indeed, NHL has been reported to occur more frequently in AIDS and in posttransplant immunocompromised patients.
8 Electron Transport Modulation
It has been shown that the desorbed intermembrane fraction of endogenous cyt-C
can function as an electron shuttle between the outer and inner membranes of intact
mitochondria. The oxidation of extramitochondrial NADH can proceed through
the rotenone-insensitive respiratory chain of the outer membrane, leading to cytb5 reduction. From cyt-b5 , electrons are channeled via the intermembrane cyt-C to
COX. In physiologic conditions, cyt-C is constitutively released outside the mitochondria to activate the exogenous NADH/cyt-C electron transport pathway. This
pathway, at least part of which routes through the VDAC, may attempt to maintain
the membrane electrochemical potential that is necessary for ATP synthesis in the
event of dysfunction of the respiratory chain upstream of complex IV [25].
It has been demonstrated that electrons from the external Nde1p have the right
of way over those coming from either Gut2p or Ndip in Saccharomyces cerevisiae.
Might a differential regulation for electron transfer in mammalian cells exist, such
that under normal conditions, electrons originating from the citric acid cycle are
preferentially used over those originating from the exogenous NADH/cyt-C pathway? Consistent with this proposition is the observation from control analysis that
NADH oxidation has negligible control over the carbon flux through OXPHOS.
Under these circumstances, blockage of the citric acid cycle or a general reduction
of mitochondrial OXPHOS could shift the bulk of the burden of electron transfer to
the exogenous NADH/cyt-C pathway. This may necessitate a greater concentration
of extramitochondrial cyt-C to maintain mitochondrial ATP levels, which, coincidently, may be enough to activate procaspase 9 and cause apoptosis. This may also
Metabolic Modulation of Carcinogenesis
111
explain why cyt-C has been chosen by biological evolution to be a mediator of both
respiration and apoptosis and why gene promoters regulating the cell cycle contain
HRE elements (see Section 11).
9 The “Switch to Glycolysis” Is Necessary to Induce Apoptosis
Lymphocytes, enterocytes, and fetal tissues that exhibit rapid growth are poorly
oxidative, whereas highly oxidative tissues such as kidney cortex or brain are normally quiescent.
It has long been noted that tumor cells display elevated rates of glycolysis,
despite oxygen availability. Crabtree observed that even though most tumor cells
are capable of oxidative phosphorylation, extracellular glucose represses oxidative phosphorylation in tumor cells via an unknown mechanism. Self-sufficiency in
growth signals and the ability to take up nutrients independently of the extracellular
environment has been postulated to be a primary acquired capability of cancer.
It has been demonstrated that eosinophils die by apoptosis even in the absence
of a functioning OXPHOS. They do so by importing and hydrolyzing glycolytic
ATP into mitochondria to maintain a transmembrane potential. This may be accomplished by operating the bidirectional Fo F1 -ATPase proton pump in reverse. Stimuli
triggering apoptosis have been shown to be partly attributable to the reverse operation of this H+ pump, which causes an increase in mitochondrial matrix pH—an
early event associated with mitochondrial dependent apoptosis.
Several reports in the literature suggest that the increase in ATP levels in cells
en route to apoptosis derives from both OXPHOS and glycolysis. It has also been
shown that apoptosis can occur in cells that lack functional mitochondria and that
ATP synthesized via glycolysis plays a critical role in preventing the collapse of
the mitochondrial membrane potential thereby allowing the cell to die by apoptosis.
Further support for this hypothesis comes from the observation that staurosporineinduced apoptotic ATP elevation in caspase-competent HeLa cells occurs via Ca2+ dependent activation of glycolytic ATP synthesis, with little or no contribution
from mitochondrial OXPHOS. TNF-induced hepatic apoptosis could be selectively
blocked by ketohexose-mediated ATP depletion.
Because apoptosis is generally recognized as an ATP-dependent process, glycolysis may provide adequate ATP to activate caspases. Procaspase 9 may be activated by
an increase in cytosolic cyt-C levels, which in turn are increased due to age-related or
pathogenic attenuation of mitochondrial OXPHOS (discussed in Section 8).
Cytosolic acidification caused by the glycolytic product, lactic acid, has been
shown to be required for induction of caspase activation and apoptosis; it also
decreases the ability of hexokinase to bind to the outer mitochondrial membrane.
Increased susceptibility to stress-induced apoptosis in c-myc transformed cells has
been linked to the overexpression of the LDH gene.
This leads to the intriguing—as yet experimentally unproven—proposition that
transient aerobic glycolysis may precede apoptosis in OXPHOS-deficient nontransformed cells and may even be necessary for such cells to undergo programmed
cell death.
112
S.P. Apte, R. Sarangarajan
On the other hand, in transformed or preneoplastic cells, where caspasedependent apoptotic pathways are blocked or rendered nonfunctional by other non–
OXPHOS-related mutations (such as PTEN, P53, TSC2, VHL, etc.), the shift to
glycolysis to supply adequate ATP to fuel caspase-dependent apoptotic pathways
may, ironically, give these cells a selective growth advantage over non-transformed
neighboring cells. Because glucose phosphorylation, the first committed step of glycolysis, is sufficient to promote cell survival by Akt, it seems possible that even a
transient shift to glycolysis in the presence of non–OXPHOS-related mutations may
promote cell immortalization.
10 Regulation of the Cell Cycle and the Apoptotic Pathway
Is Sensitive to Glucose Concentration
Increased rates of glycolysis in cultured proliferating human lymphocytes and rat
thymocytes have been reported to peak in concert with DNA synthesis. The glycolytic enzymes GAPD and LDH have been found in transcriptional coactivator
complexes with the OCT-S transcription factor, which stimulates the expression of
histone H2A as cells enter the S-phase. Similar to these glycolytic enzymes that display nonglycolytic functions among their regulatory DNA sequences, inhibition of
PFK-2, coded by the pfkfb3 gene, produces proapoptotic effects in HeLa cells. The
pfkfb3 gene promoter contains HIF-1 binding sites.
The cyclin D1 promotor contains two consensus HREs. Cyclin D1 has been
shown to be overexpressed in VHL-deficient RCC cell lines. A cyclin D1–dependent
kinase was found to phosphorylate and inactivate NRF1 with a consequent decrease
in OXPHOS. The proapoptotic protein, BAD, is required to assemble a mitochondrially located complex that also contains hexokinase IV. Furthermore, the phosphorylation status of BAD regulates hexokinase IV activity.
The GlcRE within the PK promoter was found to be sufficient to confer the modulation by HIF-1 of the glucose-dependent induction of the PK gene expression.
HIF-1 also activates telomerase via transcriptional activation of the human telomerase reverse transcriptase gene.
These results suggest that regulatory points exist in the cell cycle, oxygen sensing
and apoptotic pathways that are sensitive to glucose metabolism thereby providing
direct biochemical and genetic links between oxygen availability, metabolism, and
carcinogenesis.
11 Temporal and Spatial Effects in Tumorigenesis
In VHL-deficient RCC cell lines, the resulting stabilization of HIF-1␣ negatively
regulates mitochondrial biogenesis by inhibiting the transcription of the gene encoding the coactivator factor PGC-1␤. This is achieved by the downregulation of c-myc.
The concomitant stabilization of HIF-1␤, on the other hand, increases c-myc activ-
Metabolic Modulation of Carcinogenesis
113
ity and promotes cell-cycle progression in part through increased c-myc promoter
binding. Even though such opposing transcription factors (HIF-1␣ and HIF-1␤) may
coexist during hypoxia, the differential activation both temporally and spatially of
their downstream gene targets may still drive tumor growth instead of a net “cancellation” of their effects [26].
The variation of the intermediary metabolism of tumors with time has been proposed with respect to the cellular energy sensor, AMPK. This kinase facilitates glucose and fatty acid metabolism. However, it also switches off anabolic processes,
including proliferation through mTOR via TSC2. Active AMPK, therefore, has
properties that facilitate advanced tumorigenesis yet reduce the selective advantage
of cancer cells. The solution to this paradox is that some advanced cancers often
retain a sufficient complement of defective downstream tumor suppressors, which
allow a degree of AMPK activation while mitigating its growth-limiting effect. For
example, some cancers harbor TSC2 mutations, allowing AMPK activation without
mTOR suppression. On the other hand, for premalignant cells that are not limited
by substrate supply and that retain intact cellular senescence machinery such as p53,
TSC2, and mTOR, active AMPK is a disadvantage because it slows cellular proliferation. The inhibition of AMPK at this stage in tumorigenesis may be achieved,
in part, by increasing mitochondrial ATP production. Indeed, mitochondrial biogenesis has been shown to increase to its greatest extent at an intermediate degree of
transformation rather than at the fully transformed state.
12 The Warburg Model
The sequence of events arising from the defective functioning of the electron transfer chain in mitochondria may thus be projected to occur thus: Because the electron transfer chain uses reducing equivalents in the form of NADH as a source of
electrons, a defect in this chain (due to multiple reasons) leads to an intramitochondrial elevation of (unused) NADH levels. This leads to retrograde product inhibition
of the citric acid cycle, which in turns leads to a similar retrograde inhibition of
the NADH shuttle that transfers reducing equivalents from the cytosol to the mitochondrion [8, 27]. The increase in mitochondrial NADH levels is thus spread to the
cytoplasm. The increased NADH level in the cytosol facilitates the conversion of
pyruvate into lactate via aerobic glycolysis, thereby diverting the pool of aerobic
pyruvate away from the mitochondria to compensate for the lack of electron transport and oxidative phosphorylation. At the same time, this increased NADH/NAD+
ratio also acts to limit respiratory flux through glycolysis and promotes senescence;
both effects are due to inhibition of key glycolytic enzymes, especially GAPD.
The forced conversion of pyruvate to lactate via aerobic glycolysis [1] uses
NADH thereby serving as a mechanism to decrease NADH-mediated Akt activation [28] and [2] generates NAD+ , which can in turn activate PARP to force the cell
to undergo necrosis [29] partly by a mechanism that induces poly(ADP)ribosylation
and inactivation of GAPD [30]. An increased concentration of lactate acidifies the
114
S.P. Apte, R. Sarangarajan
cytosol, which is conducive to activation of caspases and the accumulation of the
proapoptotic protein BNIP3 [31]. Cytosolic acidification also decreases the ability
of hexokinase to bind to the outer mitochondrial membrane. 2-Methyl glyoxal, the
final product of aerobic glycolysis, acts to inhibit further increase in glycolysis by
a retrograde product inhibition. Lastly, the aerobic route of glycolysis makes the
cell more susceptible to death by glucose deprivation and death-receptor–mediated
apoptosis [32] (Table 1).
The increased cytosolic NADH levels are also postulated to increase the burden
of electron transfer to the exogenous NADH/cyt-C electron transfer pathway. This
necessitates an increase in the concentration of extra-mitochondrial cyt-C, which
coincidently may be enough to trigger activation of APAF and the activation of
procaspase 9—ultimately activating apoptosis.
Because the mitochondria cannot produce ATP from the oxidation of glucose via
the electron transfer chain, it attempts to bring in glycolytic ATP into the mitochondria via the ANT/VDAC channels. Because the import of phosphate into mitochon-
Table 1 Apoptotic Pathways Activated by Glycolysis
Attribute
Increasing
flux
through
glycolysis
Effect
NADH / NAD+
GAPD
ratio
Exogenous NADH/cyt-C
electron pathway
ROS
PARP,
NADH
Akt
NAD+
PARP
Lactate
Hexokinase binding
to mitochondrial
membrane
pH,
Caspases
Increased
apoptosis,
necrosis
GAPD
BNIP3,
PTP flux
Reverse operation of ATPase
ATP
Glycolysis, LDH
Mitochondrial pH
Disproportionate decrease in
flux through glycolysis than
through OXPHOS
Susceptibility to stress,
glucose deprivation and
death receptor
Increasing metabolic flux through glycolysis activates proapoptotic pathways and mechanisms.
These may induce death in cells whose OXPHOS is impaired but which possess functional tumor
suppressors and apoptotic machinery. Metabolic deviation may therefore act as a “self-regulating
mechanism” in such cells to prevent carcinogenesis. On the other hand, OXPHOS dysfunction
in association with other mutations may tip the balance toward proliferation, thereby causing
carcinogenesis.
Metabolic Modulation of Carcinogenesis
115
dria is blocked, the mitochondrial transmembrane potential can only be maintained
by operating the FoF1-ATPase pump in reverse, by hydrolyzing the mitochondrial
ATP pool to pump protons out into the intermembrane space. This results in the
alkalinization of the mitochondrial matrix, which, in tandem with the acidification
of the cytosol, is generally recognized as the first step to initiate apoptosis. Control
analysis studies show that ATP consumption exercises significantly more control
over the flux through glycolysis than it does through OXPHOS [33]; consequently,
an overall decreased cellular ATP level should disproportionately decrease carbon
flux through glycolysis than it does through OXPHOS.
13 Conclusion
Although aberrant functioning of the mitochondrial OXPHOS respiratory apparatus
is well documented in the literature beginning with the observations by Warburg,
emerging evidence suggests that the respiratory contribution from OXPHOS and
from the cell surface may not be negligible, when compared with that from glycolysis, in proliferating transformed cells. These observations, when taken together
with evidence suggesting that glycolysis may be a cell-death mediator of last resort
[34]—capable of activating cell death pathways either on its own or in conjunction
with OXPHOS—suggest that deficient OXPHOS may be sufficient but not necessary to initiate malignancy, and the “switch to glycolysis” may be coincidental to
but not causative of this transformation. In fact, it may be speculated that cells with
adequately functioning tumor suppressors and apoptotic machinery may be able to
derive the majority of their cellular energy requirements by “switching” back and
forth between OXPHOS and glycolysis numerous times over their life span depending on substrate or oxygen availability or as a response to moderate oxidative stress
without undergoing transformation.
In pancreatic beta cells, the ATP derived from mitochondrial pyruvate
metabolism does not significantly contribute to the glucose-stimulated insulin secretion; instead, the glycolytic metabolism of glucose is critically involved in this signaling process [35]. If the cellular energy sensor AMPK in tumor cells were altered
similarly, its “setpoint” would be decreased significantly, allowing the cancer cell to
maintain respiratory flux through anabolic metabolic processes even in the presence
of significantly reduced cellular ATP (due to impaired OXPHOS).
More studies such as the one cited in reference [36] are needed that specifically
induce a selective block of aerobic glycolysis to ascertain its effect on the initiation
and maintenance of the malignant phenotype. The contribution of pathways other
than OXPHOS and glycolysis, such as the electron transport pathway activated by
exogenous cyt-C and the trans plasma membrane electron transport pathway, should
be investigated in much more detail with regard to substrate utilization, oxygen consumption, and energy generation. Experiments should be designed to supply cell
cultures not only with just glucose or glutamine but also with the whole complement of other substrate fuels at physiologic concentrations. Metabolic alterations
116
S.P. Apte, R. Sarangarajan
should be examined in the absence of growth factor stimulation because of its facilitating effects on the metabolic shift toward aerobic glycolysis [36]. Such metabolic
alterations can be quantitatively measured through control analysis, and this technique could be expanded to include hypoxic and substrate-enhancing or -limiting
conditions as well. Levels of the whole complement of glycolytic enzymes should
be measured in order to be able to make an accurate assessment of deviations of
metabolic flux in non-transformed as well as in transformed cells. Effects should
be compared between primary, immortalized, and neoplastic cells to dissect cellspecific effects.
14 Abbreviations
AIF, apoptosis inducible factor; Akt, v-akt murine thymoma viral oncogene;
ALDH2, aldehyde dehydrogenase allele 2; AMPK, AMP activated protein kinase;
ANT, adenine nucleotide translocator; BCC, basal cell carcinoma; Bnip3, Bcl2/adenovirus E1B 19 kDa interacting protein 3; CCN, connective tissue growth
factor and nephroblastoma–overexpressed gene; CCRC, clear cell renal carcinoma; cGDPH, cytosolic glycerol-3-phosphate dehydrogenase; CHO, Chinese hamster ovary; COX, cytochrome c oxidase; CR, calorie restriction; CREB, cAMP
responsive element binding; CtBP, C-terminal binding protein; cyt-C, cytochrome
C; DM2, type 2 diabetes mellitus; E1A, adenovirus 5 early region 1A oncoprotein; ECM, extracellular matrix; eNOS, endothelial nitric oxide synthase; ES,
embryonic stem; FCCP, carbonylcyanide-p-trifluoromethoxyphenylhydrazone; FH,
fumarate hydratase; GAPD, glyceraldehyde-3-phosphate dehydrogenase; GlcRE,
glucose responsive element; GPI, glucose phosphatase isomerase; GPX1, glutathione peroxidase 1; Gut2p, glycerol 3-phosphate dehydrogenase; HBP, hexosamine biosynthetic pathway; HIF, hypoxia inducible factor; HRE, hypoxia regulating elements; IAP-2, inhibitor of apoptosis protein 2; IGF-1, insulin growth
factor; IRF3, interferon regulating factor 3; LDH, lactate dehydrogenase; LKB1, serine threonine protein kinase 11; MAVS, mitochondrial antiviral signaling protein;
Mdm2, mouse double minute 2; MDR, multidrug resistance; MEF, mouse embryo
fibroblasts; MELAS, mitochondrial encephalomyopathy, lactic acidosis and strokelike episodes; MMP, matrix metalloproteinase; mtDNA, mitochondrial DNA; mtETC,
mitochondrial electron transport chain; mTOR, mammalian target of rapamycin;
NAD+ /NADH, nicotinamide adenine dinucleotide (oxidized and reduced forms,
respectively); Nde1p and Nde2p, external NADH dehydrogenases; NDUFA1, NADH
dehydrogenase (ubiquinone) alpha subcomplex 1; NF-␬B, nuclear factor kappa B;
NHL, non–Hodgkin’s lymphoma; NQO1, NADPH dehydrogenase quinone 1; NRF,
nuclear respiratory factor; NSCLC, non–small cell lung cancer; O-GlcNAc, O-linked
acetylglucosamine; OXPHOS, oxidative phosphorylation; p21Waf1/Cip1 , cyclin dependent kinase inhibitor 1A; PDC, pyruvate decarboxylase; PDH, prolyl hydroxylase;
PDH, pyruvate dehydrogenase; PDK1, pyruvate dehydrogenase kinase 1; PFK-2,
6-phosphofructo-2 kinase/fructose-2,6-bisphosphatase; PGC, peroxisome prolifer-
Metabolic Modulation of Carcinogenesis
117
ator activator protein–␥–coactivator-1␣; PGM, phosphoglycerate mutase; PIG3, p53
inducible gene 3; PK, pyruvate kinase; PNP, purine nucleoside phosphorylase; PPAR,
peroxisome proliferator activated receptor; PTEN, phosphatase and tensin homolog;
PTP, permeability transition pore; PUFA, polyunsaturated fatty acids; RCC, renal
cell carcinoma; ROS, reactive oxygen species; SCC, squamous cell carcinoma; SDH,
succinyl dehydrogenase; SIRT1, silent information regulator 2-homolog 1; SOD,
superoxide dismutase; TCA, tricarboxylic acid; TIGAR, TP53 induced glycolysis and
apoptosis regulator; TIMP, tissue inhibitor of matrix metalloproteinase; TNF, tumor
necrosis factor; TSC2, tuberous sclerosis complex; UCP2, uncoupling protein 2;
UDP-GlcNAc, uridine diphosphate-N-acetylglucosamine; UPA, urokinase plasminogen activator; VDAC, voltage-dependent anion channel; VHL, von Hippel–Lindau;
VLB, vinblastine.
References
1. Ramanathan A, Wang C, Schreiber SL. Perturbational profiling of a cell line model of
tumorigenesis by using metabolic measurements. Proc Natl Acad Sci USA 2005; 102(17):
5992–5997.
2. Poyton RO. Assembling a time bomb – cytochrome c oxidase and disease. Nat Genet 1998;
20:316–317.
3. Pedersen P. Tumor mitochondria and the bioenergetics of cancer cells. Prog Exp Tumor Res
1978; 22:190–274.
4. Pastorino JG, Hock JB. Hexokinase II: the integration of energy metabolism and control of
apoptosis. Curr Med Chem 2003; 10(16):1535–1551.
5. Manka D, Millhorn DE. A potential molecular link between aerobic glycolysis and cancer.
Cell Cycle 2006; 5(4):343–344.
6. Lee H, Paik SG. Regulation of BNIP3 in normal and cancer cells. Mol Cells 2006; 21(1):1–6.
7. Scarpulla RC. Transcriptional activators and coactivators in the nuclear control of mitochondrial function in mammalian cells. Gene 2002; 286:81–89.
8. Matthew CK, Van Holde KE, Ahern KG. Biochemistry, 3rd ed. Boston: Addison-Wesley,
2000.
9. Baggetto LG, Lehninger AL. Isolated tumoral pyruvate dehydrogenase can synthesize acetoin
which inhibits pyruvate oxidation as well as other aldehydes. Biochem Biophys Res Commun
1987; 145(1):153–159.
10. Kagan VE, Tyurin VA, Jiang J, et al. Cytochrome C acts as a cardiolipin oxygenase required
for release of proapoptotic factors. Nat Chem Biol 2005; 1:223–232.
11. Kagan VE, Borisenko GG, Tyurina YY, et al. Oxidative lipidomics of apoptosis: redox catalytic interactions of cytochrome C with cardiolipin and phosphatidylserine. Free Radic Biol
Med 2004; 37(12):1963–1985.
12. Peskin BS, Carter MJ. Chronic cellular hypoxia as the prime cause of cancer: what is the
de-oxygenating role of adulterated and improper ratios of polyunsaturated fatty acids when
incorporated into cell membranes? Med Hypoth 2008; 70:298–304
13. Shim YJ, Choi KY, Lee WC, Kim MK, Lee SY, Lee-Kim YC. Phospholipid fatty acid patterns in the mucosa of human colorectal adenomas and carcinomas. Nutr Res 2005; 25:
261–269.
14. Baggetto LG, Testa-Parussini R. Role of acetoin on the regulation of intermediate metabolism
of Ehrlich Ascites tumor mitochondria: its contribution to membrane cholesterol enrichment
modifying passive proton permeability. Arch Biochem Biophys 1990; 283(2):241–248.
15. Bianchini F, Kaaks R, Vainio H. Overweight, obesity and cancer risk. Lancet Oncol 2002;
3(9):565–574.
118
S.P. Apte, R. Sarangarajan
16. Coughlin SS, Calle EE, Teras LR, Petrelli J, Thun MJ. Diabetes mellitus as a predictor of
cancer mortality in a large cohort of US adults. Am J Epidemiol 2004; 159:1160–1167.
17. Nisoli E, Tonello C, Cardile A, Cozzi V, Bracale R, Tedesco L. Calorie restriction promotes
mitochondrial biogenesis by inducing the expression of eNOS. Science 2005; 310(5746):
314–317.
18. Melvin RG, Ballard JWO, William O. Intraspecific variation in survival and mitochondrial oxidative phosphorylation in wild-caught Drosophila simulans. Aging Cell. 2006; 5(3):
225–233.
19. Campisi J. Senescent cells, tumor suppression and organismal aging: good citizens, bad neighbors. Cell 2005; 120:513–522.
20. Harris MH, Vander Heiden MG, Kron SJ, Thompson CB. Role of oxidative phosphorylation
in Bax toxicity. Mol Cell Biol 2000; 20(10):3590–3596.
21. Cucuianu A. Cell Darwinism, apoptosis, free radicals and haematological malignancies. Med
Hypotheses 2001; 56(1):52–57.
22. Li XD, Sun L, Seth RB, Pineda G, Chen ZJ. Hepatitis C virus protease NS3/4A cleaves mitochondrial antiviral signaling protein off the mitochondria to evade innate immunity. Proc Natl
Acad Sci USA 2005; 102(49):17717–17722.
23. Dunn GP, Old LJ, Schreiber RD. The immunobiology of cancer: immunosurveillance and
immunoediting. Immunity 2004; 21:137–148.
24. Gu Y, Wang C, Roifman CM, Cohen A. Role of MHC class I immune surveillance of mitochondrial DNA integrity. J Immunol 2003; 170:3603–3607.
25. Piana GL, Marzulli D, Gorgoglione V, Lofromento NE. Porin and cytochrome oxidase containing contact sites involved in the oxidation of cytosolic NADH. Arch Biochem Biophys
2005; 436:91–100.
26. Gordan JD, Thompson CB, Simon MC. HIF and c-myc: sibling rivals for control of cancer
cell metabolism and proliferation. Cancer Cell 2007; 12(2):108–113.
27. Nyengaard JR, Ido Y, Kilo C, Williamson JR. Interactions between hyperglycemia and
hypoxia. Diabetes 2004; 53:2931–2938.
28. Pelicano H, Xu RH, Du M, et al. Mitochondrial respiration defects in cancer cells cause
activation of Akt survival pathway through a redox–mediated mechanism. J Cell Biol 2006;
176(6):913–923.
29. Zong WX, Ditsworth D, Bauer DE, Wang ZQ, Thompson CB. Alkylating DNA damage stimulates a regulated form of necrotic cell death. Genes Dev 2004; 18:1272–1282.
30. Du X, Matsumura T, Edelstein D, et al. Inhibition of GAPDH activity by poly (ADP-ribose)
polymerase activates three major pathways of hyperglycemic damage in endothelial cells.
J Clin Invest 2003; 112: 1049–1057.
31. Kubasiak LA, Hernandez OM, Bishopric NH, Webster KA. Hypoxia and acidosis activate
cardiac myocyte death through the Bcl-2 family protein BNIP3. Proc Natl Acad Sci USA
2002; 99(20):12825–12830.
32. Pinedo CM, Ruiz CR, Almodovar CR, Palacios C, Rivas AL. Inhibition of glucose metabolism
sensitizes tumor cells to death receptor triggered apoptosis through enhancement of death
inducing signaling complex formation and apical procaspase 8 processing. J Biol Chem 2003;
278(15):12759–12768.
33. Ainscow EK, Brand MD. Internal regulation of ATP turnover, glycolysis and oxidative phosphorylation in rat hepatocytes. Eur J Biochem 1999; 266:737–749.
34. Gatenby RA, Gillies RJ. Why do cancers have high aerobic glycolysis? Cancer 2004; 4:
891–899.
35. Mertz RJ, Worley JF, Spencer B, Johnson JH, Dukes ID. Activation of stimulur-secretion
coupling in pancreatic beta cells by specific products of glucose metabolism. J Biol Chem
1996; 271(9):4838–4845.
36. Ben-Shlomo I, Kol S, Roeder LM, et al. Interleukin (IL)-1beta increases glucose uptake and
induces glycolysis in aerobically cultured rat ovarian cells: evidence that IL-1beta may mediate the gonadotropin induced midcycle metabolic shift. Endocrinol 1997; 138(7):2680–2688.
Mitochondrial DNA Mutations in Tumors
Anna Czarnecka and Ewa Bartnik
Abstract Mitochondria are subcellular organelles with the most well-known and
best-characterized function of adenosine triphosphate (ATP) production through
oxidative phosphorylation (OXPHOS). Mitochondria play an important role in
apoptosis, a fundamental biological process by which cells die in a programmed
manner and are the strategic point in the cell death cascade. Alterations in respiratory activity and mitochondrial DNA (mtDNA) abnormalities appear to be a general
feature of malignant cells. The presence of mtDNA mutations has been reported in
most types of cancer. However, the functional consequences and clinical relevance
of these mutations are not clear.
Keywords Mitochondria · Mutations · Cancer · Tumor · DNA
1 Introduction
Mitochondria are eukaryotic organelles involved in many metabolic pathways and
have a principal function of generating most of the cellular ATP through oxidative
phosphorylation (OXPHOS). Mitochondria are semiautonomous organelles that not
only perform essential functions in cellular metabolism but also play an important
role in the regulation of cell death, signaling, and free radical generation [1]. Mitochondria possess a double-membrane structure and their own genome along with
their own transcription, translation, and protein assembly machinery [2]. Nevertheless, mitochondrial functions involve an interplay between the mitochondrial and
nuclear genomes [3]. Mitochondria are called “cellular powerhouses”; the primary
function of these organelles is to produce ATP in the process of oxidative phosphorylation. At the same time, many housekeeping metabolic reactions take place in this
E. Bartnik (B)
Department of Genetics and Biotechnology, Faculty of Biology, University of Warsaw, and Institute
of Biochemistry and Biophysics, Polish Academy of sciences, Pawinskiego 5a, 02106 Warsaw,
Poland
e-mail: ebartnik@igib.uw.edu.pl
S.P. Apte, R. Sarangarajan (eds.), Cellular Respiration and Carcinogenesis,
DOI 10.1007/978-1-59745-435-3 9,
C Humana Press, a part of Springer Science+Business Media, LLC 2009
119
120
A. Czarnecka, E. Bartnik
cell compartment including production of intermediates of carbohydrate, nucleotide,
fatty acid, and amino acid metabolism. Mitochondria are indispensable for the cell
because of their key role in homeostasis, which is probably the reason for their
maintenance within cells lacking mtDNA (called ␳o ) [4]. Processes believed to be
affected by mitochondrial function include climatic adaptation, aging, longevity,
degenerative disease, and cancer [3].
Numerous diseases are known to be caused by mutations in the mtDNA. A breakthrough in mitochondrial pathophysiology occurred in 1988 when Wallace et al. [5]
found the first point mutation in the mtDNA of patients with Leber’s hereditary optic
neuropathy. To date, the etiology and pathophysiology of mitochondrial diseases
has remained perplexing. The essential role of mitochondrial oxidative phosphorylation in cellular energy production, the generation of reactive oxygen species, and
the initiation of apoptosis has suggested a number of novel mechanisms for mitochondrial pathology. Mutations in mtDNA appear to be the primary factor responsible for mitochondria-related diseases and syndromes. The number of known point
mutations, deletions, and duplications of mtDNA is more than several hundred and
new mutations are annotated every month, and the number of mtDNA polymorphisms has exceeded 1000 [6]. The recent resurgence of interest in the study of
mitochondria has been fueled in large part by the recognition that genetic and/or
metabolic alterations in this organelle are causative or contributing factors in a variety of human diseases including cancer. The role for mitochondria in tumorigenesis
was hypothesized when tumor cells were found to have an impaired respiratory
system and high glycolytic activity [7]. Recent findings elucidating the role of mitochondria in apoptosis and the high incidence of mtDNA mutations in cancer samples
further support this hypothesis [3].
Defects in mitochondrial function have long been suspected of contributing to
the development and progression of cancer. In the early 1920s, Warburg pioneered
the research on mitochondrial respiration alterations in the context of cancer and
proposed a mechanism to explain how they evolve during cell transformation. He
hypothesized that a key event in carcinogenesis involved the development of an
injury to the respiratory machinery, resulting in compensatory increases in glycolytic ATP production. Eventually, malignant cells would satisfy their energy needs
by producing a large portion of their ATP through glycolysis rather than through
oxidative phosphorylation [7]. Because of the inherent inefficiency of glycolytic
ATP generation, cancer cells represent a unique metabolic state and require high
consumption of glucose to fulfill cellular energy requirements [4]. Hypoxia itself
has been studied as a carcinogenic factor. It has been proved that aerobic glycolysis
is highly active in malignant cancers. In benign tumors, the rate of aerobic glycolysis is only one third of the malignant cancer aerobic glycolytic activity. High
aerobic glycolysis activity must correlate with glucose absorption by the cell. In
vivo measurements have been performed, and high glucose uptake by cancerous tissue in comparison with normal tissues was reported for gliomas, meningiomas, and
sarcomas [7].
To date, mitochondrial failure has been reported at all levels of their structure
and function, including abnormal ultrastructure and metabolism deregulation. First
Mitochondrial DNA Mutations in Tumors
121
of all, aerobic energy metabolism alterations have been assigned to disruption of the
nuclear encoded enzymes fumarate hydratase and succinate dehydrogenase expression (complex II of the respiratory chain). It is worth noting that mutations in these
two genes can also behave as classic tumor suppressors linking mitochondrial physiology with cancer [8]. In addition, subunits of complexes of the respiratory chain
are also encoded by the mitochondrial genome (mtDNA) [3]. Accordingly, mtDNA
alterations are frequently observed in tumor cells [4, 9]. Mutations in the mtDNA
may result in altered structure of mitochondrial proteins and thus disrupt OXPHOS,
which might shift metabolism toward anaerobic respiration. However, evidence for
direct linkage of respiratory deficiency in a specific tumor type with a specific
mtDNA mutation is still missing [8].
2 The Mitochondrial Genome
In 1981, the complete sequence of human mtDNA was determined by Sanger and
co-workers [10]. All genes have been identified and characterized [3]. Human
mtDNA is a double-stranded, closed-circular molecule of approximately 16.6 kb,
which corresponds with a molecular weight of about 10 million Da, and in most
cells it represents only about 0.5% to 1% of total DNA content [11]. The normal
mtDNA state is thought to be a supercoiled structure, and it is poorly associated
with proteins in comparison with nuclear DNA. The two strands of mtDNA can be
distinguished due to their different G content and can be separated by density in
denaturing gradients into a heavy or H-strand and a light or L-strand [2]. Each mitochondrion contains 2 to 10 copies of its genome, and a mammalian cell typically
contains 200 to 2000 mitochondria. As a consequence of mtDNA multiplication in
the cell, mitochondrial genomes can tolerate very high levels (up to 90%) of damaged DNA through complementation by the remaining wild type [2]. Mitochondrial
DNA encodes a small, but essential number of polypeptides of the OXPHOS system. The coding sequences for two ribosomal RNAs (rRNAs), 22 transfer RNAs
(tRNAs) and 13 polypeptides, which are subunits of four OXPHOS complexes,
are contiguous and lack introns (see Fig. 1 and Table 1). The tRNAs are regularly interspersed between the rRNA and protein-coding genes, playing a crucial
role in RNA maturation from the polycistronic transcripts. A single major 1.1-kblong noncoding region, called the displacement loop (D-loop), contains the main
regulatory sequences for transcription and replication initiation [2]. The D-loop is a
triple-stranded structure in which a nascent H-strand DNA segment of 500 to 700
nucleotides remains annealed to the parental L-strand. This is the region that is
most variable in sequence. However, the whole mtDNA is much more variable in
sequence than nuclear genome (nDNA), as it is more vulnerable to mutations due
to the lack of histone protection, limited repair capacity, and close proximity to the
electron transport chain, which constantly generates superoxide radicals. A higher
rate of mutation is accompanied by higher ratio of nonsynonymous to synonymous
substitution in mtDNA than in nuclear genes [12]. As mtDNA lacks introns, most
mutations occur in the coding sequences and are thus likely to be of biological
122
A. Czarnecka, E. Bartnik
Fig. 1 Human mitochondrial DNA molecule. ATP6/8, ATP synthase F0 subunit 6/8; COX I,
cytochrome c oxidase subunit I; COX III, cytochrome c oxidase subunit III; CYT B, cytochrome
b; HSP, heavy-strand promoter; LSP, light-strand promoter; ND1, NADH dehydrogenase subunit
1; ND2, NADH dehydrogenase subunit 2; ND3, NADH dehydrogenase subunit 3; ND4L, NADH
dehydrogenase subunit 4L; ND4, NADH dehydrogenase subunit 4; ND5, NADH dehydrogenase
subunit 5; ND6, NADH dehydrogenase subunit 6; OH, H-strand origin; OL, L-strand origin; A-W,
tRNAs.
consequence. The mitochondrial genome has other characteristics that distinguish
mtDNA from nDNA. These include not only a higher rate of mutation but also the
use of a divergent genetic code and transmission by maternal inheritance [3].
It has been proposed that dysregulation of mtDNA genes may contribute to
bioenergetic imbalances as it is known that some rapidly growing tumors display
alterations of oxidative metabolism and high rates of aerobic glycolysis. Further
experiments have shown that mutations in D-loop, OXPHOS genes, and mitochondrial tRNA are characteristic for cancer cells. Moreover, mutations and deletions in
mtDNA that have been observed in various solid tumors and hematologic malignancies are associated with abnormal expression of mtDNA-encoded proteins [11].
Mutations in a single mtDNA genome are silent, but as soon as the proportion of
mutant mtDNA exceeds a critical threshold concentration, a defect of OXPHOS,
integrity of inner mitochondrial membrane, electron leakage, and increased reactive
oxygen species (ROS) production result [3]. Clonal expansion of a single somatic
mtDNA mutation has substantial implications for a cell, as mtDNA mutations are at
the basis of a number of human pathologies, not only cancer. This state of art is a
derivative of the fact that a combination of different respiratory chain complexes (I,
Mitochondrial DNA Mutations in Tumors
123
Table 1 Mitochondrial Genes Encoded in the Mitochondrial Genome (mtDNA) and Nuclear
Genome (nDNA)
Mitochondrial genes
Mitochondrial complex
Genes encoded
in mtDNA
Complex I NADH
7
dehydrogenase
Complex II FAD
0
dehydrogenase
Complex III cytochrome 1
b-c1
Complex IV
3
cytochrome oxidase
Complex V ATP
2
synthase
Total
13
mtDNA genes
Total number of mtDNA
and nDNA genes
ND1, ND2, ND3, ND4,
ND4L, ND5, ND6
—
43
Cytb
11
COX1, COX2, COX3
13
ATP6, ATP8
16
4
87
III, IV, and V) defects might have substantial physiologic effects, for example may
lead to a dysregulation of oxidative phosphorylation, which in turn allows robust
production of the carcinogenic ROS. Growing molecular evidence suggests that cancer cells exhibit increased intrinsic oxidative stress, due in part to oncogenic stimulation, increased metabolic activity, and mitochondrial malfunction, which turns
on the vicious circle of oncogenesis [13]. Given the critical role of mitochondria
in apoptosis, it is conceivable that mutations in mtDNA in cancer cells could also
significantly affect the cellular apoptotic response to anticancer agents and promote
multidrug resistance [14]. At the same time, other studies have shown that mouse
cells carrying nuclei from tumors expressed tumorigenicity, whereas cell with noncancerous nuclei and mitochondria from cancer cells did not [15]. Thus at least
for some cancer cells, the mitochondria do not seem to have a profound effect on
the cancerous state of the cell. On the other hand, this does not exclude their role
in tumor growth and progression, and this has been experimentally demonstrated
[16, 17].
Interest in potential therapeutic agents that could act on the mitochondria in
tumor cells has fueled an explosion in the research on mitochondria and mtDNA
in recent years, and although the role of mitochondria in tumorigenesis remains
unclear, numerous publications emphasize the impact of mtDNA mutations in the
process of carcinogenesis [3, 4, 9]. Reports on mtDNA mutations in cancer published to date present an unclear view of overwhelming mutation/polymorphism
data with doubtful clinical correlations. Moreover, much of the literature demonstrating mutations—whether heteroplasmic or homoplasmic—has been burdened
with technical and conceptual errors that many times reduce the finding of mtDNA
mutations in tumors to the level of technical artifacts [18, 19]. Thus far, a limited number of mtDNA mutations are known that can be classified as pathologic
markers. These include a 4977-bp deletion in H¨urthle cell thyroid carcinoma;
124
A. Czarnecka, E. Bartnik
mutations of complex I and IV in thyroid carcinoma arising from A cells; and
a 264-bp deletion (3323–3588) in renal cell carcinoma. Further analysis revealed
that accumulation of mtDNA mutations is correlated with tumor aggressiveness and
multidrug resistance or with poor prognosis and acute symptoms and signs [1, 4,
11]. The increasing ease with which the mitochondrial genome can be analyzed has
helped to recognize mtDNA mutations as a frequent event in the carcinogenic process. However, it is difficult to estimate the true prevalence of mtDNA-related cancer owing to many clinical presentations and the involvement of numerous causative
mutations.
3 Mitochondrial Genome Instability in Cancer
Mitochondrial genome instability (mtGI) in cancer can be defined as the occurrence of mutations in tumor cells that are not found in the normal cells of the same
individual [20]. The mtDNA mutation rate is 100-fold higher than in the nuclear
genome [4, 11, 20], and several reasons are invoked to explain this fact, which
include (a) mtDNA is less protected by proteins, (b) it is physically associated with
the mitochondrial inner membrane where damaging ROS are generated, and (c) it
appears to have less efficient repair mechanisms than does the nucleus, as subunit ␣
of mitochondrial polymerase ␥ has only limited 3 –5 exonuclease activity [4, 20].
Also, malfunction of mismatch repair (MMR) and slipped strand mispairing (SSM)
should be considered as mtGI factors [20]. Mitochondrial DNA mutations arise not
only as a consequence of carcinogenetic stress but also as a result of polymerase
gamma errors (T to C or G to A transitions) and are accumulated in daughter cells
as consequence of genetic drift [21].
One of the intrinsic factors responsible for mtDNA mutation generation are ROS
produced by the OXPHOS system located in the inner membrane. Because the mitochondrial respiratory chain is a major source of ROS generation in the cells and the
naked mtDNA molecule is in close proximity to the source of ROS, the vulnerability
of the mtDNA to ROS-mediated damage appears to be a mechanism to amplify ROS
in already stressed cancer cells. Damaged, misfolded, or unfolded OXPHOS proteins in turn promote even more pronounced ROS production, which drives a vicious
circle of cell transformation [3, 4, 11, 20]. The proposal that mtDNA mutations and
respiratory dysfunction may be linked directly to carcinogenesis via apoptotic or
ROS-mediated pathways is challenging but urgently needs experimental proof in
cancer cells, preferably in specimens from cancer patients. To date, ROS overproduction has been demonstrated only in cells harboring mtDNA mutations responsible for mitochondrial diseases, the most common being NARP (neurogenic muscle
weakness, ataxia and retinitis pigmentosa), MELAS (mitochondrial encephalomyopathy lactic acidosis, and strokelike episodes), MERRF (myoclonic epilepsy and
ragged-red fibers), LHON (Leber hereditary optic neuropathy), and KSS (KearnsSayre syndrome) [22].
Moreover, it is worth emphasizing that some somatic mutations found in cancer
cells are sometimes identical to those found in mitochondrial disease [23–25], which
Mitochondrial DNA Mutations in Tumors
125
offers the possibility of extrapolating mitochondrial disease defects into cancer cells.
Nevertheless, further in vivo experiments are indispensable to validate the proposed
hypothesis.
4 Homoplasmy and Heteroplasmy in Cancer Cells
Cells are polyploid with respect to mtDNA: most mammalian cells contain hundreds of mitochondria and subsequently mtDNA copies. If in a given individual,
all mtDNA copies are identical, a condition known as homoplasmy arises, but if
mutations occur, are amplified and coexist with wild-type mtDNA, this is designated heteroplasmy. At cell division, mitochondria and their genomes are randomly
distributed to daughter cells, and hence, starting from a heteroplasmic situation,
homoplasmy and different levels of heteroplasmy may be found in daughter cell
lineages [2, 20, 26]. In cancer cells, mutated mtDNA can be found as homo- or
heteroplasmic [4], but many reports show that mtDNA mutations in cancer cells
are homoplasmic. This raises the question of mechanisms responsible for normal to
mutated mtDNA shift in the process of carcinogenesis [20, 26]. At least two scenarios are probable that drive this change. First of all, mutant mitochondria that lose
a certain mtDNA fragment by deletion can be considered to replicate more rapidly
than normal ones, resulting in an advantage in intracellular mitochondrial competition (replicative advantage). If the competition is intense (high rate of proliferation),
heteroplasmic cells possessing both types of mitochondria give rise to homoplasmic
daughter cells including mutant mitochondria only, with high probability. According
to mathematical models, the rate of switching from wild type to mutated mitochondrial DNA is affected by factors including the intensity of intracellular competition and the effective population size [27]. Researchers in several laboratories who
reported a high frequency of homoplasmic mitochondrial DNA mutations in human
tumors proposed not only replicative advantage for mutated mtDNA copies [28] but
also an advantage in growth and/or tumorigenic properties for a cell containing certain mtDNA mutations [24]. Polyak et al. [28] have proposed that mitochondrial
genomes replicate more rapidly to balance OXPHOS failure and reduced ATP production. On the other hand, it cannot be excluded that homoplasmy arises entirely by
chance in tumor progenitor cells, without any physiologic advantage or tumorigenic
requirement. Through extensive computer modeling, it has been demonstrated that
there is sufficient opportunity for a tumor progenitor cell to achieve homoplasmy
through unbiased mtDNA replication and sorting during cell division [29]. Computer modeling has shown that random drift is also a sufficient mechanism to explain
the homoplasmic nature of cancer cells [4]. Relaxed replication phenomenon analysis in normal cells has proved that replicative or metabolic advantage is not indispensable for homoplasmy to arise [4, 24, 26]. Although a role for other mechanisms
is not excluded, random processes are sufficient to explain the incidence of homoplasmic mtDNA mutations in human tumors. Depending on how the mutant copies
are distributed between the stem and transition cells, the mutants are depleted or
enriched, and the number of mutants in the lineage fluctuates as a random walk.
126
A. Czarnecka, E. Bartnik
On average, only approximately 70 generations are required for a mutation that is
destined to become homoplasmic. The number of 70 generations is small compared
with the number of cell divisions that a tumor progenitor cell is expected to undergo
[28]. To summarize, whereas it cannot be excluded that selection occurs in a subset
of tumors, there is no need to invoke a physiologic advantage or a role for mitochondrial mutations in tumorigenesis to explain the existence of homoplasmy or its
observed frequency [29]. However, a mathematical model cannot prove or disprove
that a phenomenon is occurring in growing cells, as it is based on assumptions and
does not constitute experimental proof.
5 Conclusion
Since the initial publications by Warburg [7] more than 50 years ago, a number of
cancer-related mitochondrial defects have been identified and described in the literature. Growing evidence suggests that cancer cells exhibit increased intrinsic mitochondrial stress, due in part to oncogenic stimulation, increased metabolic activity,
and mitochondria-nucleus signaling malfunction [30]. Despite the increased identification of signatures of mtDNA damage in transformed cells, the phenotypic effects
of these genetic changes remain to be established. While there are many reports of
these phenomena, the mechanisms responsible for the initiation and evolution of
mtDNA mutations and their roles in the development of cancer, drug resistance,
and disease progression still remain to be elucidated [9]. Research into the identification of altered expression patterns of mitochondrial proteins in cancer cells has
been made possible by the relatively recent development of mitochondrial functional proteomics. The potential of this field may be realized in the identification of new markers and risk assessment as well as therapeutic targets [14, 30].
Despite the difficulties with mitochondrial proteomics, it is likely that the combination of the mitochondrial genetic and the proteomic approaches will provide an
effective weapon in the fight against cancer. It is hoped that this strategy will provide specific genetic markers and protein profiles that will provide early detection,
risk assessment, and new targets for treatment. The recently initiated generation of
mouse models for mtDNA-linked diseases may help to solve some of those open
questions.
There are more than 400 publications in Medline reporting mitochondrial DNA
mutations in cancer. Recently, a whole issue of Oncogene was devoted to this problem. What, if any, is the take-home message about the role of mtDNA mutations in
cancer?
First of all, the literature is not entirely reliable. There are reports of data analysis that essentially denies any role for mtDNA mutations in tumors [18]. There
are problems with some of the mtDNA mutation papers, one of them being that
mtDNA haplogroups differ by specific polymorphisms throughout the mitochondrial genome, and are the results of mutations in distinct maternal lineages, as mitochondria are inherited exclusively in the maternal line [3, 31]. Thus the comparison
of any mitochondrial sequence with the Cambridge reference sequence (CRS) is
Mitochondrial DNA Mutations in Tumors
127
simply a way of determining the differences between one sequence and another—
and the results may have nothing to do with tumors and all to do with the fact that
the sample and the CRS may belong to different haplogroups of mtDNA [6]. Moreover, in most cases the analyzed groups of samples are small, for all sorts of reasons,
and conclusions cannot be drawn as to the “most common mitochondrial mutation”
in a particular type of tumor on the basis of 10 or even 20 or 30 samples as the
population of analyzed samples is not significant for statistical analysis. Thirdly,
data get misquoted; for example, a paper quotes another one out of context as to the
fact that a certain mitochondrial variant predisposes to a certain cancer even though
the original authors did not claim this and had no reason to do so. And finally, the
easily accessible mitochondrial databases are rarely used to see if something that
is suspected of being important in cancer is not simply a common polymorphism
found in many populations. The use of mtDBase [32] and MITOMAP [6] should be
obligatory for any papers dealing with mtDNA variants.
During the past 2 years, there have been increasing numbers of complete genome
sequences compared with appropriate control tissues and not with the CRS. Brandon
et al. [33] have performed a comprehensive analysis of the literature and showed
that most of the encountered mutations are population polymorphisms, and they
have put forth the attractive hypothesis that what happens in tumors is a phenomenon of selection that also took place during human adaptation to various environments [3]—though here it is the tumor cells and not the human beings who
have adapted to special circumstances. Thus most mutations in tumors would not
be dramatic changes profoundly affecting respiration, but small changes facilitating
growth under the conditions of relative hypoxia and crowding [33].
Some mutations—fairly rare—are mutations that profoundly affect cellular respiration; not many of these have been observed. Petros et al. [16] and Shidara et al.
[17] have shown that tumors containing mitochondrial DNA with the pathogenic
NARP mutation at position A8993G in ATP6 grow faster and are more resistant to
apoptosis than are tumors with the same type of cells but with wild-type mtDNA.
This was accepted as evidence that the severe mitochondrial mutations in tumors
may play a role of facilitating growth and decreasing susceptibility to apoptosis
under the exacting conditions of tumor growth. However, Salas et al. [18] have suggested that as this particular mutation has not been found in tumors, this experiment
does not prove anything. On the other hand, the data presented there are quite convincing as the growth/apoptosis perturbations can be abrogated by expression of an
unmutated gene (expressed from a gene introduced into the nucleus, as it is not yet
possible to introduce genes into human mitochondria) [17]. Brandon et al. [33] have
suggested that such severe mutations arise in early stages of tumor development and
are successively lost. Recently, Gasparre et al. [34] have shown that in oncocytic
tumor, such mutations (defined as disruptive; i.e., nonsense or frameshift) occur in
about one quarter of the tumors, and they all occur in complex I genes. However—
and this would be in agreement with Brandon et al. [33]—cell cultures grown from
these tumors lose the severe mutations [34].
Zhou et al. [35] have demonstrated for the first time that a ND2 (again complex I) mutation found in a tumor (human squamous cancer of the head and neck)
128
A. Czarnecka, E. Bartnik
can increase growth of cells in culture, thus showing that growth can depend on
the expression of a mutated mitochondrial gene. In this case, they also used a
gene introduced into the nucleus. Moreover, they showed that ROS production was
increased [35].
Thus there certainly are mutations in mtDNA in cancer cells and they do affect
the growth properties of these cells. It should be borne in mind, however, that the
mutations that are observed in most tumors are likely to be ones that have only
slight adaptive effects on their growth, and the dramatic mutations are rare. We
certainly have spent a lot of time looking for them and essentially found only
one [23].
Another problem important both for scientists and for epidemiologists is the
question whether some of us are more predisposed to certain types of tumors. Mitochondrial DNAs form haplogroups, and the differences between them believed until
recently to be completely neutral and insignificant may contribute to the frequency
of some diseases. Again there is a problem with the data. Statistics is a doubleedged sword; if the sample sizes are too small, many associations are uncertain.
Samuels et al. [31] have recently shown that large cohorts are required for studies
of associations of diseases with mitochondrial haplogroups. Studies with 500 controls and 500 cases would not always reliably show a relative risk of about 1.75 for
a common haplogroup; larger sizes would be required for less common ones [31].
Rarely are the analyzed groups even one tenth of this size. Thus far, two studies of
cancer—haplogroup associations seem quite robust—a study by Booker et al. [36]
showing increased relative risk for haplogroup U for prostate cancer and renal carcinoma (though this has again been criticized by Salas et al. [18]), and a study of the
G10398A polymorphism in ND3 in haplogroup N has shown that it is associated
with an increased breast cancer risk [37].
Increased relative risks for certain haplogroups and certain types of tumors might
be of interest to health care providers as predictors of cancer development. What
could the rationale be for this? Wallace [3] has postulated that the differences in
mitochondrial DNA sequence transferred to the function of the electron transport
chain in mitochondria could give important differences in, for example, ROS production, which could affect how susceptible the cells would be to damage. The
main problem with this hypothesis is that no differences have thus far been shown
between various normal types of cells in ROS production for humans, though a number of pathogenic mutations have been shown to affect ROS production. Recently,
Moreno-Loshuertos et al. [38] have shown differences between various mouse mitochondrial haplotypes both in respiration and in free radical production, though this
has aroused some discussion [39].
As the techniques of mutation analysis improve and more data become available,
the role of mitochondrial mutations and haplogroup associations should be elucidated in the near future. However, they are unlikely to be specific markers for any
type of tumor, and there will probably not be very strong associations between haplogroups and tumor occurrence. Finally, the importance of the highest standards in
mtDNA analysis to produce reliable results in the many investigations in carcinogenesis should be emphasized [18, 19].
Mitochondrial DNA Mutations in Tumors
129
Acknowledgment This work was partly supported by an intramural grant from The Faculty of
Biology, University of Warsaw and grant NN 401 2327 33 from the Ministry of Science and Higher
Education.
References
1. Amuthan G, Biswas G, Zhang SY, Klein-Szanto A, Vijayasarathy C, Avadhani NG.
Mitochondria-to-nucleus stress signaling induces phenotypic changes, tumor progression and
cell invasion. EMBO J 2000; 20:1910–1912.
2. Fernandez-Silva P, Enriquez JA, Montoya J. Replication and transcription of mammalian
mitochondrial DNA. Exp Physiol 2003; 88:41–56.
3. Wallace DC. A mitochondrial paradigm of metabolic and degenerative diseases, aging, and
cancer: a dawn for evolutionary medicine. Annu Rev Genet 2005; 39:359–407.
4. Carew JS, Huang P. Mitochondrial defects in cancer. Mol Cancer 2002; 1:1–12.
5. Wallace DC, Singh G, Lott MT, et al. Mitochondrial DNA mutation associated with Leber s
hereditary optic neuropathy. Science 1988; 242:1427–1430.
6. Ruiz-Pesini E, Lott MT, et al. An enhanced MITOMAP with a global mtDNA mutational
phylogeny. Nucleic Acids Res 2007; 35:D823–D828.
7. Warburg O. On the origin of cancer cells. Science 1956; 123:309–314.
8. Gottlieb E, Tomlinson JP. Mitochondrial tumor suppressors: a genetic and biochemical update.
Nat Rev Cancer 2005; 5:857–866.
9. Czarnecka AM, Golik P, Bartnik E. Mitochondrial DNA mutations in human neoplasia. J Appl
Genet 2006; 47:67–78.
10. Anderson S, Bankier AT, Barrell BG, et al. Sequence and organization of the human mitochondrial genome. Nature 1981; 290:457–465.
11. Penta JS, Johnson FM, Wachsman JT, Copeland WC. Mitochondrial DNA in human malignancy. Mutat Res 2001; 488:119–133.
12. Lynch M. Mutation accumulation in transfer RNAs: molecular evidence for Muller’s ratchet
in mitochondrial genomes. Mol Biol Evol 1996; 13:209–220.
13. Birch-Machin M. Using mitochondrial DNA as a biosensor of early cancer development. Br
J Cancer 2005; 93:271–272.
14. Don AS, Hogg PJ. Mitochondria as cancer drug targets. Trends Mol Med 2004; 10:372–378.
15. Akimoto M, Niikura M, Ichikawa M, et al. Nuclear DNA but not mtDNA controls tumor
phenotypes in mouse cells. Biochem Biophys Res Commun 2005; 327:1028–1035.
16. Petros JA, Baumann AK, Ruiz-Pesini E, et al. MtDNA mutations increase tumorigenicity in
prostate cancer. Proc Natl Acad Sci USA 2005; 102:719–724.
17. Shidara Y, Yamagata K, Kanamori T, et al. Positive contribution of pathogenic mutations in
the mitochondrial genome to the promotion of cancer by prevention from apoptosis. Cancer
Res 2005; 65:1655–1663.
18. Salas A, Yao YG, Macaulay V, Vega A, Carracedo A, Bandelt HJ. A critical reassessment of
the role of mitochondria in tumorigenesis. PLoS Med 2005 2:e296.
19. Zanssen S, Schon EA. Mitochondrial DNA mutations in cancer. PLoS Med 2005; 2:
1082–1084.
20. Bianchi NO, Bianchi MS, Richard SM. Mitochondrial genome instability in human cancer.
Rev Mut Res 2001; 488:9–23.
21. Brown DT, Samuels DC, Michael EM, Turnbull DM, Chinnery PF. Random genetic drift
determines the level of mutant mtDNA in human primary oocytes. Am J Hum Genet 2001;
68:533–536.
22. Kirkinezos IG, Moraes CT. Reactive oxygen species and mitochondrial diseases. Semin Cell
Dev Biol 2001; 12:449–457.
23. Lorenc A, Bryk J, Golik P,et al. Homoplasmic MELAS A3243G mtDNA mutation in colon
cancer sample. Mitochondrion 2003; 3:119–124.
130
A. Czarnecka, E. Bartnik
24. Chinnery PF, Samuels DC, Elson J, Turnbull DM. Accumulation of mitochondrial DNA mutations in ageing, cancer, and mitochondrial disease: is there a common mechanism? Lancet
2002; 360:1323–1325.
25. Meierhofer D, Mayr JA, Fink K, Schmeller N, Kofler B, Sperl W. Mitochondrial DNA mutations in renal cell carcinomas revealed no general impact on energy metabolism. Br J Cancer
2006; 94:268–274.
26. Augenlicht LH, Heerdt BG. Mitochondria: integrators in tumorigenesis? Nat Genet 2001;
28:104–105.
27. Yamauchi A. Rate of gene transfer from mitochondria to nucleus: effects of cytoplasmic inheritance system and intensity of intracellular competition. Genetics 2005; 171:1387–1396.
28. Polyak K, Li Y, Zhu H, et al. Somatic mutations of the mitochondrial genome in human
colorectal tumors. Nat Genet 1998; 20:291–293.
29. Coller HA, Khrapko K, Bodyak ND, Nekhaeva E, Herrero-Jimenez P, Thilly WG. High frequency of homoplasmic mitochondrial DNA mutations in human tumors can be explained
without selection. Nat Genet 2001; 28:147–150.
30. Pelicano H, Carney D, Huang P. ROS stress in cancer cells and therapeutic implications. Drug
Resist Updat 2004; 7:97–110.
31. Samuels DC, Carothers AD, Horton R, Chinnery PF. The power to detect disease associations
with mitochondrial DNA haplogroups. Am J Hum Genet 2006; 78:713–720.
32. Ingman M, Gyllensten U. mtDB: human mitochondrial genome database, a resource for population genetics and medical sciences. Nucl Acids Res 2006; 34:D749–D751.
33. Brandon M, Baldi P, Wallace DC. Mitochondrial mutations in cancer. Oncogene 2006;
25:4647–4662.
34. Gasparre G, Porcelli AM, Bonora E, et al. Disruptive mitochondrial DNA mutations in complex I subunits are markers of oncocytic phenotype in thyroid tumors. Proc Natl Acad Sci
USA 2007; 104:9001–9006.
35. Zhou S, Kachhap S, Sun W, et al. Frequency and phenotypic implications of mitochondrial
DNA mutations in human squamous cell cancers of the head and neck. Proc Natl Acad Sci
USA 2007; 104:7540–7546.
36. Booker LM, Habermacher GM, Jessie BC, et al. North American white mitochondrial haplogroups in prostate and renal cancer. J Urol 2006; 175:468–472; discussion 472–473.
37. Darvishi K, Sharma S, Bhat AK, Rai E, Bamezai RN. Mitochondrial DNA G10398A polymorphism imparts maternal Haplogroup N a risk for breast and esophageal cancer. Cancer
Lett 2007; 249:249–255.
38. Moreno-Loshuertos R, Acin-Perez R, Fernandez-Silva P, et al. Differences in reactive oxygen species production explain the phenotypes associated with common mouse mitochondrial
DNA variants. Nat Genet 2006; 38:1261–1268.
39. Battersby BJ, Shoubridge E. Reactive oxygen species and the segregation of mtDNA sequence
variants. Nat Genet 2007;39:571–572.
Cellular Respiration and Tumor Suppressor
Genes
Luis F. Gonzalez-Cuyar, Fabio Tavora, Iusta Caminha, George Perry,
Mark A. Smith, and Rudy J. Castellani
Abstract More than 70 years have passed since Dr. Otto Warburg first documented
that cancer cells relied primarily on aerobic glycolysis. However, it was not until
the late-1980s and 1990s when cellular respiration, cellular oxygen sensors, and
hypoxia were convincingly related to tumorigenesis and tumor progression. With
the discovery of hypoxia inducible factor (HIF-1) and its target genes, the relationship that was once weak has become very strong. The expression of the HIF-1
molecule has been proved to occur by hypoxic stress and confers an adaptation
advantage, upregulating genes involved in angiogenesis and glycolysis among others. It has been reported that HIF-1 is under the control of tumor suppressor genes
such as the von Hippel–Lindau (pVHL) protein, which, under normoxia, ubiquitinates HIF-1 for proteosomal degradation. Additionally, the tumor suppressor gene
p53 has been reported to be stabilized by HIF-1 and inhibits Mdm2-dependent
degradation. Recently, it was reported that synthesis of cytochrome c oxidase, a
regulator of the cytochrome c oxidase pathway, is a p53 target. Moreover, interactions of pVHL and p53 stabilize p53 and permits p53-mediated apoptosis. The aim
of this chapter is to outline how the cellular microenvironment affects cellular respiration and perhaps induces tumorigenesis, as well as to discuss ways in which tumor
suppressor genes are also involved.
Keywords Cellular respiration · Hypoxia inducible factor · p53 · Tumor suppressor
· Von Hippel–Lindau
1 Introduction
For more than 70 years, the notion that cancer cells rely preferentially on glycolysis, a phenomenon identified as the Warburg effect [1], has been in existence, but
only over the past 15 or so years have major advances been made in elucidating
R.J. Castellani (B)
Department of Pathology, School of Medicine, University of Maryland, 22 South Greene Street,
Baltimore, MD, 21201
e-mail: rcastellani@som.umaryland.edu
S.P. Apte, R. Sarangarajan (eds.), Cellular Respiration and Carcinogenesis,
DOI 10.1007/978-1-59745-435-3 10,
C Humana Press, a part of Springer Science+Business Media, LLC 2009
131
132
L.F. Gonzalez-Cuyar et al.
the role of oxygen and cellular respiration in neoplasia. A number of specific and
well-characterized genes are now known to be involved, directly and indirectly,
as well as a number of different well-characterized molecular cascades. Prominent among these is hypoxia inducible transcription factor (HIF-1), which upregulates the transcription of several target genes involved in cellular adaptation to the
tumoral microenvironment (i.e., hypoglycemia, acidosis, and low oxygen tensions)
and angiogenesis. Additionally, HIF-1 is under the posttranscriptional control of
pVHL that ubiquitinates HIF-1 and tags it for proteosomal degradation.
Two tumor suppressor genes have been identified in connection with changes in
the respiratory metabolism in cancer: TP53 and the VHL tumor suppressor gene.
TP53 is the most common mutated gene in human tumors. Whereas the majority
of the studies on p53 in past decades have dealt with its role in cell division and
death/apoptosis, some evidence now points toward its involvement also with cellular metabolism. Matoba et al. [2] recently found that the gene that encodes synthesis of cytochrome c oxidase 2 (SCO2) is upregulated in a p53-dependent manner.
SCO2 may be a transcriptional target for p53 [3], and this relation may explain
how inactivation of p53, as seen with somatic and familial mutations, may promote oncogenesis by decreasing cellular dependence on oxygen [2]. In fact, other
studies have linked p53 in the control of cell glycolysis by influence of downstream molecules. For example, Kondoh et al. [4] showed that phosphoglycerate
mutase, another glycolytic enzyme, is also controlled by p53 by posttranscriptional
activity. Activation of the Akt kinase pathway is influenced by p53 and may also
play a role in switching the cell metabolism from oxidative phosphorylation to
glycolysis [5].
The pVHL protein is a positive regulator of p53, and germ-line or somatic mutations on the VHL gene are related to tumors such as renal cell carcinomas, hemangioblastomas, and pheochromocytomas. It has been postulated that VHL protein
interacts with HIF-1 tagging it for proteosomal degradation [6]. In the absence of
oxygen under normal circumstances or when pVHL is mutated, the HIF-1 transcription factor is inactivated with direct action upon molecules such as vascular
endothelial growth factor (VEGF), culminating in angiogenesis and, thus, cancer
progression [7].
In this chapter, we will discuss the role of tumor suppressor genes in the
metabolism of cells, more specifically in the various components of cellular respiration, interactions with HIF-1, and how interactions of the gene products will
promote oncogenesis, cell survival, and cancer progression through upregulation
of transcription of genes involved in angiogenesis, resistance to apoptosis, and
metastasis.
2 Otto Warburg and Cellular Respiration
Otto Heinrich Warburg (1883–1970), a German biochemist, made major contributions to the study of respiration and neoplasia. He hypothesized early on that as cancer cells have an abnormal growth rate, the means by which the required energy is
Respiration and Tumor Suppressor Genes
133
obtained may be central to the pathogenesis of neoplasia [1, 8]. Results from earlier
works demonstrating that rates of respiration of the sea urchin egg after fertilization increase up to sixfold motivated him to study cancer cells in a similar manner
[1, 8]. As he measured the rates of respiration in transplantable cancer cells from the
Flexner-Jobling rat carcinoma, he concluded that there was no increase in oxygen
consumption and lactate was being produced by glycolysis, even in the presence of
oxygen [1, 8]. Warburg concluded that contrary to cells in normal tissues, cancer
cells could not suppress glycolysis in the presence of oxygen. Interestingly, he also
described aerobic glycolysis in tissues exposed to ethylcarbilamine.
Warburg further demonstrated that although there was no significant increment in
levels of cellular respiration or ATP production of cancer cells, anaerobic glycolysis
provided a major energy source. He hypothesized that when normal cells become
cancerous that metabolic activity shifts in favor of glycolysis. This switch in his
opinion was secondary to a “dedifferentiation” or “reorientation of gene expression”
that diverted the baseline energy production to growth.
After a paper by Goldblatt and Cameron [9] citing induction of fibrosarcoma
from fibroblasts exposed to intermittent hypoxia, he demonstrated that embryonic cells exposed to hypoxia reverted to glycolysis and that normal metabolism,
as in cancer, was irreversible, even in the presence of oxygen [1, 8]. Additionally he pointed out that subsequent animal inoculation led to the development of
cancers.
In 1956, Warburg described that irreversible damage to the cellular respiration
can be achieved by low oxygen tensions, respiratory poisons such as arsenic, hydrogen sulfide, and urethane, and x-rays [1, 8].
In an excerpt from a 1967 paper entitled “The prime cause and prevention of
cancer,” he simplified his conclusions:
Cancer, above all other diseases, has countless secondary causes. Almost anything can cause
cancer. But, even for cancer, there is only one prime cause. Summarized in a few words, the
prime cause of cancer is the replacement of the respiration of oxygen in normal body cells
by a fermentation of sugar.
Although his work was not studied for close to 70 years, Warburg’s conclusions
on the rates of respiration are still valid with respect to tumorigenesis research and
cancer therapeutics.
3 Molecular Homeostasis and Genetic Instability: Genetic
Instability and Tumorigenesis
There is a consensus pertaining to the biological definition of life, which is summarized in seven categories: homeostasis, organization, metabolism, growth, adaptation, response to stimuli, and reproduction. Mammals, as the highest life form,
have an exquisite intrinsic balance of all of these functions at the molecular level
[10–12]. Hanahan and Weinberg [13] described six essential characteristics dictating tumorigenesis and progression, which are somewhat analogous to the above:
134
L.F. Gonzalez-Cuyar et al.
self-sufficiency of growth signals, insensitivity to growth-inhibitory signals, evasion
of programmed cell death, limitless replicative potential, sustained angiogenesis,
and tissue invasion and metastases. Huang et al. [14] proposed that tumor development is composed of expansion and progression preceded by genetic adaptation
(glycolysis, angiogenesis) and genetic alteration (malignant selection of characteristics), respectively.
Decades of scientific research in the area of tumorigenesis and neoplasia have
yielded an insurmountable amount of information as to how all of these essential characteristics of life are deranged in neoplasia [14–16]. Current thought suggests that cancer is a genetic disease that originates as a stepwise process in which
normal cells go through several premalignant phenotypes by acquiring cumulative
genetic alterations each conferring a more malignant phenotype leading eventually to progression and metastases [13]. Alterations in the genetic composition of
cells either by nucleotide mutations or chromosomal aberrations are regarded as
the initiator of neoplasia, while accumulation of these lead to progression, invasion, and metastases [14, 17]. The process by which clones are selected thus follows selection for mutations that confer a survival advantage in a hostile tumoral
microenvironment [14, 17].
A central question with respect to oncogenesis is that of etiology. Although this
remains an open question, current thought generally suggests that adverse conditions in the tumor microenvironment contribute to genetic instability and induction
of mutagenesis by direct DNA damage and damage to DNA repair systems [18, 19].
According to some, intermittent hypoxia/reperfusion leads to generation of reactive oxygen species (ROS), leading in turn to the formation of 8-hydroxyguanine
and thymine glycols, which lead to transversions [18–20]. Apart from such oxidative damage, aberrant DNA synthesis as seen by amplification as well as increased
incidence of point mutations contribute to further progression [18, 19].
4 Hypoxia
Hypoxia is among the most studied stressors in solid tumors, the central theme being
the shift from aerobic respiration to glycolysis due to limited availability of oxygen
and nutrients. Hypoxia occurs in tumor tissue that is between 100 and 200 ␮m from
a functional blood supply [17, 21, 22]. Normal tissue displays an oxygen diffusion
gradient of 400 ␮m from a blood supply, but significant hypoxia is seen in tumor
cells even directly adjacent to a capillary with a mean oxygen tension of 2%, and
cells located 200 ␮m from capillaries showed 0.02% [22, 23]. In addition, studies
have suggested that hypoxia is an independent prognostic indicator.
Reynolds et al. [20] demonstrated a fivefold increase in mutations in both tumors
and cells cultured in hypoxia. Hypoxia has also been implicated as a stimulus for
p53 activation and accumulation [24]. Hypoxia seems to not only induce the genetic
instability that accounts for cumulative genetic aberration but also selects for malignant clones. It has been suggested that hypoxia-induced apoptosis can be mediated
Respiration and Tumor Suppressor Genes
135
by HIF-1 upregulation of p53-induced apoptosis. Therefore, to an extent, HIF-1
selects for tumor clones that have lost p53 and thus are less sensitive to hypoxia and
hypoglycemia.
5 HRE and HIF-1
In 1992, Semenza and Wang [12] described a 50-nt region located 116-nt 3 of the
polyadenylation site at the flanking sequence of the human erythropoietin (EPO)
gene that was able to dramatically increase transcription in the presence of low
oxygen tensions. The region known as the hypoxia responsive element (HRE)
increased the transcription of EPO up to 7-fold in cells cultured at 2% O2 and
up to 20-fold in cells cultured at 1% O2 [12]. Also it was described that mutations within the 4–12 nt and 19–23 nt would not produce the desired increase in
transcription, thus it was hypothesized that these sites were crucial for binding of
transcription factors [12]. Concordantly with manipulation of these mutation sites,
researchers were able to discover HIF-1, which binds to the HRE at the 1–18 nt
region, thus increasing the transcription of the desired gene under hypoxic but
not normoxic conditions [12]. Because maximal levels of EPO mRNA after the
induction by hypoxia takes about 2 to 4 hours, this suggested that maximization of
gene products involving HRE in the face of hypoxia required de novo synthesis of
proteins.
Further molecular characterization of HIF-1 revealed that it is a heterodimeric
transcription factor composed of two subunits, HIF-1␣ and HIF-1␤, which is also
known as the aryl hydrocarbon nuclear translocator (ARNT) [11, 12, 25–28]. These
proteins have a basic helix-loop-helix (bHLH) configuration with a Per, Ahr/ARNT
domain (PAS) [11, 12, 25–28]. Whereas the former facilitates the nuclear dimerization of the two subunits, the latter is required for DNA binding [25, 26, 29, 30].
Thus far, three isoforms have been described for each subunit, HIF-1, 2, 3␣, and
ARNT1, ARNT2, ARNT3, respectively [28, 29, 31–33]. It is important to note that
the ARNT is constitutively transcribed, whereas the transcription of the ␣ subunit
is stimulated by low oxygen tensions as well as several oncogenes [29, 31]. When
dimerized, the subunits bind to the cis-acting HRE in the RCGTG conserved motif
mediated by Ser22, Ala25, Arg30 [28, 29, 31–33]. Subsequently, increased transcription of the gene products involved in cellular adaptations to hypoxia such as
VEGF and EPO are attained [12, 28, 29, 31–35].
The N-terminal region of the HIF-1␣ molecule is necessary for dimerization,
whereas the C-terminus is required for nuclear localization, transactivation, and
stabilization of the protein. The molecule also contains an oxygen dependent
degradation (ODD) domain in the region of residues 401–603, which in addition
contains two proline residues, 401 and 564, which are essential for degradation
under normoxic conditions [31]. There are two areas of nuclear localization signals
(NLS), one on the N-terminal (residues 17–33) and one on the C-terminal (residues
718–721) [31, 36]. Additionally, HIF-1␣ contains two transactivation domains
136
L.F. Gonzalez-Cuyar et al.
(TAD) at the C-terminal 531–575 and 786–826 [31, 33]. The area between these
two nucleotides contains a domain that inhibits transactivation.
Approximately 53% of malignant tumors express HIF-1 by immunohistochemistry including colon, breast, gliomas, prostate, renal cell carcinomas, and glioblastomas when compared with the respective normal tissue. Low-grade gliomas have
little or no staining with HIF-1. Glioblastomas show intense staining, particularly
in the pseudopalisading cells and in the neighboring areas of necrosis. Apparently,
malignant gliomas take advantage of the HIF-1 activation by upregulation of platelet
derived growth factor (PDGF) and VEGF, which appear directly proportional. To
date, approximately 30 target genes that are transactivated by HIF-1 have been identified as downstream targets of HIF-1 including genes involved in glycolysis, glucose transport, angiogenesis, vascular remodeling and tone, erythropoiesis, and iron
metabolism [10, 15].
6 Oxygen Sensors
Under normoxic conditions, both subunits of HIF-1, HIF-1␣ and HIF-1␤, are transcribed into primary transcripts and exported to the cytoplasm where they are translated. Whereas constitutively expressed HIF-1␤ is stable under both normoxic and
hypoxic conditions, HIF-1␣ is unstable under normoxic conditions. In the presence
of oxygen, it is hydroxylated at two proline residues located on amino acid 401 and
amino acid 564, and then tagged by pVHL, forming part of the ubiquinone-ligase
complex [25, 29, 37–40], rendering it susceptible to proteosomal degradation. It is
important to note that these proline hydroxylases that account for the protein’s posttranscriptional modification are oxygen dependent, therefore under hypoxic conditions the proline residues will not be hydroxylated so the protein will be translocated
to the nucleus where it would dimerize with its counterpart [25, 29, 37–40]. After
dimerization, the molecule interacts with several HREs located upstream of gene
products involved in cellular adaptations to hypoxia, acidosis, and hypoglycemia.
It has also been suggested that under hypoxic conditions, the pVHL–HIF-1
complex still forms but with ineffective degradation of HIF-1␣ [41]. Also, in the
presence of iron or desferrioxamine, the pVHL–HIF-1 complex is not formed,
thus demonstrating another way of HIF-1 stabilization [42]. However, it has been
reported that this type of induction can be reversed by cyclohexamide, showing the
need of de novo protein synthesis [43]. Additionally, in wild-type pVHL cell lines,
investigators described that the HIF-1 DNA binding doublet contains pVHL, thus
pointing out that there is pVHL–HIF-1 complex formation under normoxia [41].
7 Tumor Microenvironment
The tumor microenvironment is fundamentally hostile for normal cellular
metabolism, being characterized by hypoxia, hypoglycemia, and acidosis, thereby
enhancing the selection process for malignant clones. In this regard, it is not
Respiration and Tumor Suppressor Genes
137
surprising that hypoxia has been established as an independent adverse prognostic factor for a variety of tumor types. Moreover, aberrant tumor vasculature may
contribute to the low oxygen tension, and this lack of adequate perfusion, coupled
with cellular overpopulation, leads to decreased cellular “resources” and increased
need for the cells to adapt.
As mentioned above, the hostile microenvironment at the epicenter of the tumor
affects a shift from oxidative metabolism to glycolytic pathways. HREs have been
identified in a variety of genes involved in this shift. These include EPO and VEGF,
as well as several of the enzymes required for glycolysis such as glyceraldehyde
phosphate dehydrogenase, enolase, phosphofructokinase, phosphoglycerate kinase,
and pyruvate kinase, as well as other genes such as transferrin, transferrin receptor,
IGF binding proteins, lactate dehydrogenase, and glucose transporters (GLT-1 and
GLT-3).
Some authors have hypothesized that products of glycolytic metabolism, such as
pyruvate, can further stabilize the HIF-1␣ molecule by substituting 2-oxoglutarate
from the proline hydroxylase and thus inhibit the hydroxylation of HIF-1␣ at the
ODD domain [38]. Both lactate and carbonic anhydrases 9 and 12 may further
foster an acidic environment contributing to tumorigenesis. These carbonic anhydrases appear to be also responsive to HIF-1–HRE induction, as they are not
normally expressed in normoxic cells. Moreover, tumors that have a faulty glycolytic pathway, secondary to increased mutations, still have an acidic environment [44], which further suggests the role of these anhydrases in promoting a tumor
microenvironment.
8 Cellular Respiration and Tumor Suppressor Genes
8.1 p53
p53 is a tumor suppressor gene located on 17p13.1 and encodes a 393-amino-acid
DNA- binding protein that is mutated in more than 50% of human tumors and
that is responsible for inhibiting uncontrolled cellular proliferation and invasion
[2, 45]. p53 induces cell cycle arrest and apoptosis by enhanced transcription of
target genes involved in every step of the apoptosis signaling pathway in response
to DNA damage by various stimuli [3, 45–49]. p53 is activated by hypoxia, DNA
damage, expression of certain oncogenes, and cytotoxic stimuli and, as with all
tumor suppressor genes, inactivation is accomplished by loss in both alleles, usually by missense mutations [50, 51]. In the absence of cellular stress, the p53 gene
product is ubiquitinated and tagged for proteosomal degradation by the ubiquitin
ligase Mdm2 [24, 45]. Graber et al. [52] also described that the tumor expressing
wild-type p53 underwent apoptosis in hypoxic areas whereas the mutant p53 cells
would show resistance to hypoxia-induced apoptosis, giving the p53-deficient cells
a selective advantage over their wild-type p53 counterparts [24, 52]. Under hypoxia
and addition of HIF-1 mimetics such as CoCl2 and DFX [53], it has been demonstrated that HIF-1 and p53 associate leading to p53 stabilization [24, 54] Moreover,
138
L.F. Gonzalez-Cuyar et al.
it was reported by Carmeliet et al. [55] that accumulation of HIF-1 was induced
by anoxia and hypoglycemia only in the presence of HIF-1. However, others have
reported that accumulation of p53 is induced by the usual HIF-1 stimuli, even in
cells that are HIF-1 deficient [24, 56].
p53 is the most common mutated gene in cancer cells, and it has been demonstrated that it can also affect metabolism [2, 3]. p53 has been demonstrated to have
a role in cellular respiration, and it has a direct impact in cellular respiration by
contributing to the assembly of cytochrome c oxidase complex [2]. In mice, loss
of p53 resulted in decreased oxygen consumption, increased lactic acid production,
but comparable ATP production [2]. However, p53-dependent mice had a marked
decreased in stamina, which suggests that p53 plays a crucial role in adequate ATP
production in sustained physical exercise [2,3]. SCO2 is a transcription target of
p53, and p53-deficient cells have decreased levels of SCO2 and no aerobic respiration. After reintroduction of SCO2 in p53 null cells, cells are then driven into
glycolysis due to increased ATP requirements. Matoba et al. [2] described SCO2
as the first direct target that modulates energy metabolism [2, 3]. SCO2 is required
for adequate assembly of a critical component of the respiratory chain cytochrome
c oxidase. Thus, the loss of p53 and the subsequent decrease in SCO2 might be
the switch that leads cells to use a glycolytic pathway therefore giving the cell the
capability to adapt to increased energy requirements.
p53-SCO2 connects tumor suppressors to cellular respiration and to metabolism,
which correlates with the previous views on how cellular respiration and tumorigenesis intersect. Mutations in p53 render cells less sensitive to apoptosis induced
by hypoxic states, and tumors that lack p53 appear to be more resistant to radiation and chemotherapy [49]. Approximately 90% of mutations in p53 are point
mutations of residues affecting the DNA binding domain [49, 57–59]. Additionally, p53 posttranscriptionally controls the level of phosphoglycerate mutase
(PGM) [4].
Mechanistically, there are many schools of thought. For example, some find that
hypoxia induces the accumulation of p53 not the downstream upregulation of its
target genes [60]. On the other hand, others suggest that, under prolonged hypoxia,
HIF-1 and p53 form a complex with Mdm2 therefore inhibiting HIF-1 downstream
gene activation [61].
8.2 von Hippel–Lindau Syndrome
The von Hippel–Lindau syndrome takes its name from German ophthalmologist
Eugen von Hippel and Swedish neuropathologist Arvid Lindau who, in the early
19th century, described a syndrome of angiomas in the retina and the cerebellum,
respectively. The VHL gene is located at 3p16-15 and is ubiquitously expressed.
It consists of three exons that are transcribed into two transcripts that in turn are
translated into three proteins [62]. The first transcript contains all three exons and
Respiration and Tumor Suppressor Genes
139
is translated into two proteins, VHL30 and VHL19 containing 213 and 160 amino
acids, respectively, both of which are active in tumor suppression [63]. The second
transcript contains exons 1 and 3 and is hypothesized to have decreased tumor suppressor activity [64–66]. The gene product consists of a 30-kDa protein that has an
␣ subunit, comprising residues 155–213, and a ␤ subunit that includes the first 154
residues [67, 68].
VHL exon 2 encodes for the ␤ domain, and exon 3 encodes for the ␣ domain
[69] The ␤ domain is necessary for conjugation with HIF-1 and fibronectin and
apparently mediates nuclear cytoplasmic shuttling of pVHL independent of both
O2 and HIF-1. The ␣ domain is encoded by exon 3 and is required for the formation of the BC–cullin-2 complex, E3 ligase activity. Several mutations of the
VHL gene account for three main phenotypes of this familial syndrome. The most
common phenotype includes cerebellar hemangioblastomas and clear cell renal cell
carcinoma with or without adrenal pheochromocytomas. It is often considered an
autosomal dominant disease because although both VHL copies need to be mutated
according to the Knudson two-hit hypothesis, with one inherited mutant copy of
VHL, the second mutation is almost guaranteed [64, 70–72]
Previously, we mentioned that pVHL interacts with HIF-1␣ under normoxia after
hydroxylation of two proline residues of the latter by oxygen-dependent proline
hydroxylases, which targets the protein for proteosomal degradation [16, 29, 71,
73, 74]. Elongin C recruits an E3 ubiquitin ligase complex that in turn consists of an
Elongin B, CUL2, and RBX1. Additionally, an E2 ubiquitin-conjugating enzyme
attaches to this complex and, later, mediates the ubiquitination of HIF-1. The ␤
subunit will recognize HIF-1-OH.
Under hypoxic conditions, the hydroxylation of HIF-1 and pVHL-ligase complex
recognition does not take place, and this leads to dimerization and nuclear translocation of both HIF-1 subunits with subsequent binding to HREs and transcriptional
upregulation of target genes. Taking into consideration that HIF-1 mediates angiogenesis by transcriptional upregulation of vascular endothelial growth factor receptor (VGFR) among others and that pVHL is necessary for HIF-1 degradation, it is
no surprise that two carcinomas with known VHL mutations, such as CCRCC and
cerebellar hemangioblastomas, have such a prominent and characteristic vascular
network [41, 72]. Several studies point to the fact that pVHL inactivation has only
been linked to renal cell carcinoma (RCC) and hemangioblastoma (HB). In these
tumors, the cellular composition varies from VHL–/– to VHL+/− whereas endothelial cells are VHL+/− and VHL−/− [72].
Immunocytochemistry shows cellular and vascular proliferation seen by the
autocrine-acting TGF-␣ and paracrine-acting VEGF, PDGF, respectively, both
downstream genes upregulated by HIF-1 [72]. Furthermore, it has been suggested
that the pVHL has more targets, for example the pheochromocytoma-only variant (2C), which might retain its ability to ubiquitinate HIF-1 [37, 67, 68, 71, 72,
75, 76]. Not surprisingly, tumors notorious for their prominent vasculature are part
of the VHL syndrome, perhaps due to ineffective ubiquitination of HIF-1␣ leading
to upregulation of VEGF and other angiogneic gene products.
140
L.F. Gonzalez-Cuyar et al.
pVHL has also been reported to control the extracellular matrix (ECM) by controlling fibronectin deposition [77, 78]. In pVHL-deficient cells, the extracellular
deposition seen possibly contributes to increased migration and perhaps increases
their metastatic potential [78]. Fibronectin 1 expression is not affected by hypoxia,
suggesting a HFI-1–independent mode of regulation by pVHL [62]. Thus, there are
tumor suppressor pathways of pVHL that are independent of HIF-1 [62, 79, 80].
Additionally, it has been described that pVHL also controls ECM degradation regulating metalloproteinases 2 and 9, their inhibitors, and the urokinase type plasminogen activator system [62, 81–83].
Other genes regulated by the HIF-1/pVHL complex include aminopetidase A,
collagen type IV, alfa1, cyclin G2, DEC1/Straqw, endothelin 1, low-density lipoprotein receptor-related protein-1, and MIC/CD99 [79, 80]. In fact, the total number of
genes that pVHL targets, either directly or independently of HIF-1, is likely to be
of the order of hundreds [80]. Mutations in VHL will stabilize HIF-1 and contribute
to tumorigenesis and progression [69]. Independently of HIF-1, through inadequate
deposition of ECM proteins, mutations in VHL might select for tumor clones with
a phenotype that can more readily invade and metastasize.
8.3 Interactions Between p53 and VHL
Apart from their independent contributions to tumor suppression, it has recently
been described that the gene products of both genes can also work with each other
for their common goal. In 2006, Roe and colleagues [84] showed that p53 binds to
pVHL at the latter’s ␣ domain and thus competes for this site with Elongin C. As we
mentioned before, the domain of pVHL is the one responsible for the recognition
and subsequent proteosomal degradation of HIF-1␣. The pVHL-p53 complex leads
inhibition of the Mdm2-mediated ubiquitination of p53, consequently leading to p53
stabilization and upregulated downstream transcription of proapoptotic p21 and Bax
promoters [62]. Additionally, it has been shown that it increases the acetylation and
association with p300 under genotoxic stress and thus p53 transcriptional activity
and p53 mediated cell cycle arrest [62, 84, 85].
9 Conclusion
Conditions in the cellular microenvironment, such as hypoxia, not only lead to
genetic instability but also lead to the activation of molecular mechanisms determined to confer a survival advantage by phenotypic adaptation. Mechanisms such
as the HIF-1 cascade permit the cells to attain these capabilities by activating downstream genes. Limited data exists on how the tumor suppressor genes potentially
affect cellular respiration. This aspect perhaps will be a potential target for cancer
therapy in the future.
Respiration and Tumor Suppressor Genes
141
References
1. Warburg O. On the origin of cancer cells. Science 1956; 123:309–314.
2. Matoba S, Kang JG, Patino WD, et al. p53 regulates mitochondrial respiration. Science 2006;
312:1650–1653.
3. Kruse JP, Gu W. p53 aerobics: the major tumor suppressor fuels your workout. Cell Metab
2006; 4:1–3.
4. Kondoh H, Lleonart ME, Bernard D, Gil J. Protection from oxidative stress by enhanced
glycolysis: a possible mechanism of cellular immortalization. Histol Histopathol 2007; 22:
85–90.
5. Plas DR, Thompson CB. Akt-dependent transformation: there is more to growth than just
surviving. Oncogene 2005; 24:7435–7442.
6. Mahon PC, Hirota K, Semenza GL. FIH-1: a novel protein that interacts with HIF-1alpha and
VHL to mediate repression of HIF-1 transcriptional activity. Genes Dev 2001; 15:2675–2686.
7. Semenza GL. Targeting HIF-1 for cancer therapy. Nat Rev Cancer 2003; 3:721–732.
8. Krebs H. Otto Warburg: Cell physiologist, biochemist and eccentric. Oxford: Oxford
University Press, 1981.
9. Goldblatt H, Cameron G. Induced malignancy in cells from rat myocardium subjected to
intermittent anaerobiosis during long propagation in vitro. J Exp Med 1953; 97:525–552.
10. Semenza GL. Hypoxia, clonal selection, and the role of HIF-1 in tumor progression. Crit Rev
Biochem Mol Biol 2000; 35:71–103.
11. Semenza GL, Roth PH, Fang HM, Wang GL. Transcriptional regulation of genes encoding
glycolytic enzymes by hypoxia-inducible factor 1. J Biol Chem 1994; 269:23757–23763.
12. Semenza GL, Wang GL. A nuclear factor induced by hypoxia via de novo protein synthesis
binds to the human erythropoietin gene enhancer at a site required for transcriptional activation. Mol Cell Biol 1992; 12:5447–5454.
13. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000; 100:57–70.
14. Huang LE, Bindra RS, Glazer PM, Harris AL. Hypoxia-induced genetic instability—a calculated mechanism underlying tumor progression. J Mol Med 2007; 85:139–148.
15. Semenza GL. Regulation of physiological responses to continuous and intermittent hypoxia
by hypoxia-inducible factor 1. Exp Physiol 2006; 91:803–806.
16. Semenza GL. Hypoxia and cancer. Cancer Metastasis Rev 2007; 26(2):223–234.
17. Dang CV, Semenza GL. Oncogenic alterations of metabolism. Trends Biochem Sci 1999;
24:68–72.
18. Bindra RS, Crosby ME, Glazer PM. Regulation of DNA repair in hypoxic cancer cells. Cancer
Metastasis Rev 2007; 26:249–260.
19. Bindra RS, Glazer PM. Genetic instability and the tumor microenvironment: towards the concept of microenvironment-induced mutagenesis. Mutat Res 2005; 569:75–85.
20. Reynolds TY, Rockwell S, Glazer PM. Genetic instability induced by the tumor microenvironment. Cancer Res 1996; 56:5754–5757.
21. Helmlinger G, Yuan F, Dellian M, Jain RK. Interstitial pH and pO2 gradients in solid tumors
in vivo: high-resolution measurements reveal a lack of correlation. Nat Med 1997; 3:177–182.
22. Folkman J. Incipient angiogenesis. J Natl Cancer Inst 2000; 92:94–95.
23. Denko NC, Fontana LA, Hudson KM, et al. Investigating hypoxic tumor physiology through
gene expression patterns. Oncogene 2003; 22:5907–5914.
24. Pan Y, Oprysko PR, Asham AM, Koch CJ, Simon MC. p53 cannot be induced by hypoxia
alone but responds to the hypoxic microenvironment. Oncogene 2004; 23:4975–4983.
25. Wang GL, Jiang BH, Rue EA, Semenza GL. Hypoxia-inducible factor 1 is a basic-helixloop-helix-PAS heterodimer regulated by cellular O2 tension. Proc Natl Acad Sci USA 1995;
92:5510–5514.
26. Wang GL, Semenza GL. General involvement of hypoxia-inducible factor 1 in transcriptional
response to hypoxia. Proc Natl Acad Sci USA 1993; 90:4304–4308.
142
L.F. Gonzalez-Cuyar et al.
27. Wang GL, Semenza GL. Purification and characterization of hypoxia-inducible factor 1. J Biol
Chem 1995; 270:1230–1237.
28. Michel G, Minet E, Ernest I, et al. A model for the complex between the hypoxia-inducible
factor-1 (HIF-1) and its consensus DNA sequence. J Biomol Struct Dyn 2000; 18:169–179.
29. Acker T, Plate KH. A role for hypoxia and hypoxia-inducible transcription factors in tumor
physiology. J Mol Med 2002; 80:562–575.
30. Semenza GL, Nejfelt MK, Chi SM, Antonarakis SE. Hypoxia-inducible nuclear factors bind
to an enhancer element located 3 to the human erythropoietin gene. Proc Natl Acad Sci USA
1991; 88:5680–5684.
31. Shi YH, Fang WG. Hypoxia-inducible factor-1 in tumour angiogenesis. World J Gastroenterol
2004; 10:1082–1087.
32. Jiang BH, Rue E, Wang GL, Roe R, Semenza GL. Dimerization, DNA binding, and transactivation properties of hypoxia-inducible factor 1. J Biol Chem 1996; 271:17771–17778.
33. Jiang BH, Zheng JZ, Leung SW, Roe R, Semenza GL. Transactivation and inhibitory domains
of hypoxia-inducible factor 1 alpha: modulation of transcriptional activity by oxygen tension.
J Biol Chem 1997; 272:19253–19260.
34. Kourembanas S, Hannan RL, Faller DV. Oxygen tension regulates the expression of the
platelet-derived growth factor-B chain gene in human endothelial cells. J Clin Invest 1990;
86:670–674.
35. Kourembanas S, Marsden PA, McQuillan LP, Faller DV. Hypoxia induces endothelin gene
expression and secretion in cultured human endothelium. J Clin Invest 1991; 88:1054–1057.
36. Vandromme M, Gauthier-Rouviere C, Lamb N, Fernandez A. Regulation of transcription factor localization: fine-tuning of gene expression. Trends Biochem Sci 1996; 21:59–64.
37. Clifford SC, Astuti D, Hooper L, Maxwell PH, Ratcliffe PJ, Maher ER. The pVHL-associated
SCF ubiquitin ligase complex: molecular genetic analysis of elongin B and C, Rbx1 and HIF1alpha in renal cell carcinoma. Oncogene 2001; 20:5067–5074.
38. Lu H, Forbes RA, Verma A. Hypoxia-inducible factor 1 activation by aerobic glycolysis implicates the Warburg effect in carcinogenesis. J Biol Chem 2002; 277:23111–23115.
39. Mabjeesh NJ, Amir S. Hypoxia-inducible factor (HIF) in human tumorigenesis. Histol
Histopathol 2007; 22:559–572.
40. Semenza GL. HIF-1 mediates the Warburg effect in clear cell renal carcinoma. J Bioenerg
Biomembr 2007; 39(3):231–234.
41. Maxwell PH, Wiesener MS, Chang GW, et al. The tumour suppressor protein VHL targets
hypoxia-inducible factors for oxygen-dependent proteolysis. Nature 1999; 399:271–275.
42. Goldberg MA, Dunning SP, Bunn HF. Regulation of the erythropoietin gene: evidence that
the oxygen sensor is a heme protein. Science 1988; 242:1412–1415.
43. Wang GL, Semenza GL. Desferrioxamine induces erythropoietin gene expression and
hypoxia-inducible factor 1 DNA-binding activity: implications for models of hypoxia signal
transduction. Blood 1993; 82:3610–3615.
44. Wykoff CC, Beasley NJ, Watson PH, et al. Hypoxia-inducible expression of tumor-associated
carbonic anhydrases. Cancer Res 2000; 60:7075–7083.
45. Levine AJ. p53, the cellular gatekeeper for growth and division. Cell 1997; 88:323–331.
46. Freedman DA, Epstein CB, Roth JC, Levine AJ. A genetic approach to mapping the p53
binding site in the MDM2 protein. Mol Med 1997; 3:248–259.
47. Wu L, Levine AJ. Differential regulation of the p21/WAF-1 and mdm2 genes after high-dose
UV irradiation: p53-dependent and p53-independent regulation of the mdm2 gene. Mol Med
1997; 3:441–451.
48. Hollstein M, Rice K, Greenblatt MS, et al. Database of p53 gene somatic mutations in human
tumors and cell lines. Nucleic Acids Res 1994; 22:3551–3555.
49. Michalak E, Villunger A, Erlacher M, Strasser A. Death squads enlisted by the tumour suppressor p53. Biochem Biophys Res Commun 2005; 331:786–798.
50. Kinzler KW, Vogelstein B. Life (and death) in a malignant tumour. Nature 1996; 379:19–20.
Respiration and Tumor Suppressor Genes
143
51. Vogelstein B, Kinzler KW. Cancer genes and the pathways they control. Nat Med 2004;
10:789–799.
52. Graeber TG, Osmanian C, Jacks T, et al. Hypoxia-mediated selection of cells with diminished
apoptotic potential in solid tumours. Nature 1996; 379:88–91.
53. Garber K. Energy boost: the Warburg effect returns in a new theory of cancer. J Natl Cancer
Inst 2004; 96:1805–1806.
54. Ravi R, Mookerjee B, Bhujwalla ZM, et al. Regulation of tumor angiogenesis by p53-induced
degradation of hypoxia-inducible factor 1alpha. Genes Dev 2000; 14:34–44.
55. Carmeliet P, Dor Y, Herbert JM, et al. Role of HIF-1alpha in hypoxia-mediated apoptosis, cell
proliferation and tumour angiogenesis. Nature 1998; 394:485–490.
56. Wenger RH, Camenisch G, Desbaillets I, Chilov D, Gassmann M. Up-regulation of hypoxiainducible factor-1alpha is not sufficient for hypoxic/anoxic p53 induction. Cancer Res 1998;
58:5678–5680.
57. Vousden KH. Activation of the p53 tumor suppressor protein. Biochim Biophys Acta 2002;
1602:47–59.
58. Ryan KM, Vousden KH. Cancer: pinning a change on p53. Nature 2002; 419:795–797.
59. Vousden KH, Lu X. Live or let die: the cell s response to p53. Nat Rev Cancer 2002; 2:
594–604.
60. Koumenis C, Alarcon R, Hammond E, et al. Regulation of p53 by hypoxia: dissociation of
transcriptional repression and apoptosis from p53-dependent transactivation. Mol Cell Biol
2001; 21:1297–1310.
61. Schmid T, Zhou J, Kohl R, Brune B. p300 relieves p53-evoked transcriptional repression of
hypoxia-inducible factor-1 (HIF-1). Biochem J 2004; 380:289–295.
62. Ohh M. Ubiquitin pathway in VHL cancer syndrome. Neoplasia 2006; 8:623–629.
63. Latif F, Tory K, Gnarra J, et al. Identification of the von Hippel-Lindau disease tumor suppressor gene. Science 1993; 260:1317–1320.
64. Gnarra JR, Tory K, Weng Y, et al. Mutations of the VHL tumour suppressor gene in renal
carcinoma. Nat Genet 1994; 7:85–90.
65. Herman JG, Latif F, Weng Y, et al. Silencing of the VHL tumor-suppressor gene by DNA
methylation in renal carcinoma. Proc Natl Acad Sci USA 1994; 91:9700–9704.
66. Sekido Y, Bader S, Latif F, et al. Molecular analysis of the von Hippel-Lindau disease tumor
suppressor gene in human lung cancer cell lines. Oncogene 1994; 9:1599–1604.
67. Kaelin WG Jr. The von Hippel-Lindau tumor suppressor gene and kidney cancer. Clin Cancer
Res 2004; 10:6290S–6295S.
68. Kim WY, Kaelin WG. Role of VHL gene mutation in human cancer. J Clin Oncol 2004;
22:4991–5004.
69. Bonicalzi ME, Groulx I, de Paulsen N, Lee S. Role of exon 2-encoded beta-domain of the von
Hippel-Lindau tumor suppressor protein. J Biol Chem 2001; 276:1407–1416.
70. Knudson AG Jr. Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci
USA 1971; 68:820–823.
71. Semenza GL. VHL and p53: tumor suppressors team up to prevent cancer. Mol Cell 2006;
22:437–439.
72. Kondo K, Klco J, Nakamura E, Lechpammer M, Kaelin WG Jr. Inhibition of HIF is necessary
for tumor suppression by the von Hippel-Lindau protein. Cancer Cell 2002; 1:237–246.
73. Semenza GL. HIF-1: using two hands to flip the angiogenic switch. Cancer Metastasis Rev
2000; 19:59–65.
74. Semenza GL. Oxygen-dependent regulation of mitochondrial respiration by hypoxiainducible factor 1. Biochem J 2007; 405:1–9.
75. Clifford SC, Cockman ME, Smallwood AC, et al. Contrasting effects on HIF-1alpha regulation by disease-causing pVHL mutations correlate with patterns of tumourigenesis in von
Hippel-Lindau disease. Hum Mol Genet 2001; 10:1029–1038.
76. Hoffman MA, Ohh M, Yang H, Klco JM, Ivan M, Kaelin WG Jr. von Hippel-Lindau protein
mutants linked to type 2C VHL disease preserve the ability to downregulate HIF. Hum Mol
Genet 2001; 10:1019–1027.
144
L.F. Gonzalez-Cuyar et al.
77. Bluyssen HA, Lolkema MP, van Beest M, et al. Fibronectin is a hypoxia-independent target
of the tumor suppressor VHL. FEBS Lett 2004; 556:137–142.
78. Ohh M, Yauch RL, Lonergan KM, et al. The von Hippel-Lindau tumor suppressor protein
is required for proper assembly of an extracellular fibronectin matrix. Mol Cell 1998; 1:
959–968.
79. Xu Q, Briggs J, Park S, et al. Targeting Stat3 blocks both HIF-1 and VEGF expression induced
by multiple oncogenic growth signaling pathways. Oncogene 2005; 24:5552–5560.
80. Wykoff CC, Pugh CW, Maxwell PH, Harris AL, Ratcliffe PJ. Identification of novel hypoxia
dependent and independent target genes of the von Hippel-Lindau (VHL) tumour suppressor
by mRNA differential expression profiling. Oncogene 2000; 19:6297–6305.
81. Merzak A, Koocheckpour S, Pilkington GJ. CD44 mediates human glioma cell adhesion and
invasion in vitro. Cancer Res 1994; 54:3988–3992.
82. Los M, Zeamari S, Foekens JA, Gebbink MF, Voest EE. Regulation of the urokinase-type
plasminogen activator system by the von Hippel-Lindau tumor suppressor gene. Cancer Res
1999; 59:4440–4445.
83. Zatyka M, Morrissey C, Kuzmin I, et al. Genetic and functional analysis of the von HippelLindau (VHL) tumour suppressor gene promoter. J Med Genet 2002; 39:463–472.
84. Roe JS, Kim H, Lee SM, Kim ST, Cho EJ, Youn HD. p53 stabilization and transactivation by
a von Hippel-Lindau protein. Mol Cell 2006; 22:395–405.
85. Sansone P, Storci G, Pandolfi S, Montanaro L, Chieco P, Bonafe M. The p53 codon 72 proline allele is endowed with enhanced cell-death inducing potential in cancer cells exposed to
hypoxia. Br J Cancer 2007; 96:1302–1308.
Uncoupling Cellular Respiration: A Link to
Cancer Cell Metabolism and Immune Privilege
M. Karen Newell, Elizabeth M. Villalobos-Menuey, Marilyn Burnett,
and Robert E. Camley
Abstract Epidemiologic evidence strongly correlates suppression of maturation of
the immune system to incidences of cancer. The effectiveness of chemotherapeutic
agents significantly depends upon the ability of the cancer cells to express (costimulatory) molecules on their surface that can be recognized and engaged by their
surrogate ligands of the immune system, thereby resulting in immune-directed cell
killing. Alterations in the metabolic strategies of cancer cells (such as: the rate of
glucose utilization, glucose or fatty acid oxidation, the magnitude of the membrane
potential and the pH gradient across the mitochondrial membrane, and/or changes in
uncoupling protein levels) can modulate the expression of some cell surface “death
receptors”; thereby conferring resistance to certain cell death-inducing stimuli.
Keywords Cancer cell metabolism · Fas · Drug resistance · Uncoupling protein ·
Metabolic strategy · Metabolic modulators · Mitochondria · Cellular respiration ·
Immune privilege
1 History
Cancer cells exhibit distinct metabolic pathways to meet their energy demands. Otto
Warburg postulated that there must be some sort of deficit or abnormality in what
has long been described as the most energy-efficient strategy that a cell can use—
mitochondrial respiration. The purpose of this chapter is to propose an alternate
explanation and to present a hypothesis that can be rigorously tested.
M.K. Newell (B)
CU Institute of Bioenergetics and Immunology, University of Colorado at Colorado Springs,
Colorado Springs, CO 80918
e-mail: mnewell@uccs.edu
e-mail: rcastellani@som.umaryland.edu
S.P. Apte, R. Sarangarajan (eds.), Cellular Respiration and Carcinogenesis,
DOI 10.1007/978-1-59745-435-3 11,
C Humana Press, a part of Springer Science+Business Media, LLC 2009
145
146
M.K. Newell et al.
2 An Overview of Cellular Metabolism
Every cell in the body uses carbohydrates, protein, or fat in different proportions
to ensure that the cell has sufficient energy to perform its normal function. The
cell’s choice of fuel (i.e., the cell’s metabolic strategy) changes depending on its
activation or differentiation state as well as its environment. For example, a cell
that is dividing has different energy demands than does one that is nondividing and,
thus, may employ an alternative metabolic strategy. Another example would be the
change in strategy for a cell that has been damaged by infection or stress.
The majority of nondividing cells in the body use carbons derived from carbohydrate or glucose to efficiently produce the cell’s ultimate energy unit of exchange,
adenosine triphosphate (ATP). The process begins with the binding of glucose to
the glucose receptor/transporter and transport into the cytosol, where the molecule
becomes phosphorylated by hexokinase. Once phosphorylated, the molecule continues to be processed into two molecules of pyruvate. Pyruvate then enters the Krebs
cycle, a process that occurs in the mitochondria where ultimately the production of
NADH and FADH2 provide reducing equivalents that flow through the four complexes of the electron transport chain, eventually creating a proton gradient between
the inner mitochondrial membrane and the matrix. In concert, the electrochemical
gradient drives the synthesis of adenosine diphosphate (ADP) to ATP as a result of
the electrochemical gradient across the inner mitochondrial membrane.
3 Tumor-Specific Metabolic Pathways
Rapidly dividing tumor cells display a characteristic metabolic strategy that distinguishes the tumor cell from its normal, nondividing counterpart cell of the tissues. The key difference is that the rate of glycolysis is much greater than the
rate of glucose breakdown in a normal cell under resting conditions. Glycolysis
is an oxygen-independent process of splitting the six-carbon glucose molecules and
thereby providing energy in the form of ATP to the cell. Aerobic cells use the glycolysis pathway as the initial pathway that ultimately involves oxygen consumption
and complete oxidation of the carbons derived from carbohydrates. From a series of
10 reactions, the first five of which are energy “investments” and the second five of
which are energy generating, the final product is pyruvate [1].
The fate of the pyruvate molecules produced by glycolysis largely depends on the
oxidative state of the cell in question. Glycolysis needs a pool of cytosolic NAD+ ,
derived from the oxidation of NADH, to proceed. In the presence of oxygen (aerobic
glycolysis), the NADH is oxidized in the mitochondria. If the rate of glycolysis
in the cell is faster than the rate at which the NADH can be oxidized by electron
transport in the mitochondria, the excess NADH must then be used to drive the
reduction of another substrate to maintain cellular redox balance. Pyruvate itself is
such a substrate, and by the activity of the enzyme lactate dehydrogenase, pyruvate
is reduced to lactic acid [1].
Uncoupling Cellular Respiration
147
Glycolysis, whether aerobic or anaerobic, releases only a very small percent of
the stored energy in the glucose molecule. The evolution of aerobic metabolism
made possible a more energy-efficient use of the stored energy in glucose and in
other carbon-containing fuels. Pasteur was the first to observe that when cells metabolizing glucose anaerobically were exposed to oxygen, the rate of glycolysis slowed
dramatically (a phenomenon known as the Pasteur effect). This effect likely occurs
because it is far more energy efficient to completely oxidize glucose than via just
the initial glycolytic steps alone. Much of the control over these processes is determined, not surprisingly, by the energy needs of the cell [1].
4 Mitochondrial Metabolism
Pyruvate derived from carbohydrate is one of the three suppliers of acetyl-CoA
for oxidation in the Krebs cycle. The conversion of pyruvate to acetyl-CoA is catalyzed by the enzyme pyruvate dehydrogenase. The glycolytic product, pyruvate, is
transported from the cytosol to the mitochondria. Once there, the Ca2+ -dependent
enzyme pyruvate dehydrogenase catalyzes its production to acetyl-CoA and in that
manner feeds the glucose-derived carbons to the Krebs cycle ultimately providing
the reducing equivalents necessary for electron transport, oxidative phosphorylation, and the very efficient enzymatic production of ATP. The activity of pyruvate
dehydrogenase requires a high mitochondrial membrane potential across the inner
mitochondrial membrane [1].
When pyruvate supplies are high or when there is insufficient Ca2+ to activate pyruvate dehydrogenase, the source of the acetyl-CoA switches from glucosederived pyruvate to the acetyl-CoA produced by fatty acid oxidation. The source of
fatty acid can be endogenous or exogenous, composed of short-, medium-, or longchain fatty acids. When long-chain fatty acids are used, the process that provides
acetyl-CoA is beta-oxidation. A key enzyme in the transport of fatty acids is carnitine palmitoyl transferase (either I or II depending on the tissue of origin). Tumor
cells in general and drug-resistant tumor cells in particular, as well as other cells that
remain “immune privileged,” have a flexible capacity for the ready switch to fatty
acid oxidation as a result of mitochondrial expression of key players, including mitochondrial uncoupling proteins that uncouple glycolysis from oxidative phosphorylation. This process results in a lower mitochondrial membrane potential, decreased
sensitivity to free radical damage, and insensitivity to apoptosis inducing stimuli,
such as chemotherapy or radiation [1].
5 The Immune Response in Cancer
Our work has addressed the possibility that the immune system monitors the
metabolic state and degree of damage that characterizes all cells. We have found
that, though perhaps preferentially destroying those in an excessively damaged state
(i.e., stressed/damaged cells or infected cells), nonetheless, tumor cells often survive
this surveillance by changing their metabolic strategy to one that is protective [24].
148
M.K. Newell et al.
When tumor cells are treated with traditional chemotherapeutic drugs or radiation,
the damage from these treatments causes cell surface changes allowing the cell to
be recognized once again by the immune system. Drug-resistant cell variants display a characteristic metabolic state that masks them from the immune system (even
after treatment with chemotherapeutic drugs or radiation) and that confers upon the
tumor cell profound protection from further insults [2].
We have focused on understanding the characteristic metabolic states that protect
abnormally proliferating cells. Cancer cells, particularly multidrug-resistant variants, have a strategy for survival that involves selective fatty acid oxidation in the
mitochondria. We have identified several drug candidates that block the activity
of these proteins. Preliminary studies indicate that these compounds are capable
of interfering with the metabolic strategy of most tumor cells and have striking
therapeutic activity in tumor-bearing mice when used alone or in conjunction with
standard chemotherapy. We are currently investigating several novel drugs designed
to target metabolic strategies used by tumor cells, in particular multidrug-resistant
tumors, and other diseases characterized by abnormally proliferating cells.
Attempts to trigger an effective immune response against growing cancer cells
have been a priority in a large body of research. This is not surprising given the
results that tumor allografts are routinely rejected by an effective immune response,
the observation that children and the elderly (considered somewhat immunocompromised) have increased frequencies of tumors relative to the general population,
and the increased incidence of tumors in immunocompromised hosts. The problems
have been to understand enough about how the immune system detects abnormal
proliferation and, likely more importantly, how an immune response can distinguish
self from non-self, leading to the question of whether tumor cells are foreign or self.
Clearly, in many instances there are associations with viral infections, although in
other cases there are no clear links between a pathogen and tumorigenesis.
6 Fundamental Questions
Every year at least 6.2 million people die worldwide from cancer [3]. The development of drug resistance attenuates the effectiveness of chemotherapy in many cases.
In this chapter, we explore whether chemotherapy can be enhanced by increasing
the response of the immune system to the tumor cells and by making drug-resistant
tumor cells more visible to the immune system.
Specifically, we ask the following questions:
1. Do effective chemotherapeutics sensitize tumors to immune destruction by
increasing cell surface expression of costimulatory molecules?
2. Can this effect be enhanced by other treatments?
3. Is drug resistance, in part, due to differences in metabolic behavior between drugsensitive and drug-resistant cells? Using this idea, can we create new methods for
treating drug-resistant tumors?
Uncoupling Cellular Respiration
149
7 The Immune System and Cancer
Extraordinary advances in our understanding of the immune system have been made
in the past 100 years, especially since the discovery of the T cell and the B cell
[4–6]. Na¨ıve T cells require two signals for activation. These are recognition of
antigens in major histocompatibility complex (MHC)-encoded molecules [4] and
a costimulation signal [7–11] provided by the B7/CD28 family members or other
costimulatory molecules such as Fas (CD95) [12]. Previously activated T cells can
be reactivated by costimulation alone [13, 14]. In the absence of activation, T cells
disregard the tissue. If a T cell is activated, the consequences can be (1) destruction
of the damaged cells or (2) repair of damaged cells by promoting regeneration either
directly or indirectly.
There is substantial evidence that the immune system plays an extensive role in
suppressing cancer. However, it is also clear that the ability of the immune system
to control cancer is not perfect.
Researchers have tried to stimulate the immune system as a therapeutic strategy
against cancer for many years [15, 16]. These attempts have generally been ineffective, although there have been some recent successes [17]. There are many reasons
for this variability. These include (1) the inability to activate T cells that can destroy
the tumor due to the absence of signal one (recognition of the appropriate tumor
antigen); (2) the presence of a signal two that results in the production of the wrong
cytokines by T cells, which may lead to the growth of tumors, or (3) the failure of
activated T cells to kill cancerous cells [18].
In contrast with therapies that concentrate primarily on activating the immune
system, we will explore the idea that effective chemotherapeutics work in concert
with the immune system and derive some of their effectiveness from the immune
response. This means that by tuning the immune response, we may be able to
increase effectiveness in cancer therapy.
Cellular metabolism may play a dominant role in modulating the intercellular communication between a cancer cell and a T cell and thereby plays a dominant role in
T-cell activation. In particular, we have observed a significant correlation between
cellular metabolism and the expression of cell-surface Fas, one of the key costimulatory molecules. Fas is expressed on most rapidly dividing and self-renewing cells,
including a wide variety of tumor cells [19]. Whereas Fas is widely known as a “death
receptor” and can induce programmed cell death when engaged by Fas ligand (FasL,
CD95L) or anti-Fas antibodies [20, 21], paradoxically, Fas engagement can also result
in accelerated proliferation of T lymphocytes and some tumor cells [22, 23].
Cell surface Fas expression is upregulated in response to increasing glucose concentrations in vitro in transformed cell lines and in freshly isolated cells from a
variety of tissue origins [19]. Obviously, glucose is a key source of fuel, and the
availability of glucose drives metabolic activity. A further example of the connection between cellular metabolism and Fas is found in the work of Bhushan et al. [2]
who showed that drug-resistant tumor cells fail to express cell surface Fas and the
work of Harper et al. [24] who showed that drug-resistant tumor cells also demonstrated a unique metabolic strategy, quite different from drug-sensitive tumor cells.
150
M.K. Newell et al.
8 Drug Treatment and Fas-Induced Tumor Cell Death
Anticancer agents may work to promote the death of tumor cells in multiple ways.
First, these agents may work by direct cytolysis requiring active participation of
the tumor cell in the death process [25]. Second, chemotherapeutics may promote
the ability of the tumor cell to be recognized by cells of the immune system and
to be killed by immune-directed cell death [26]. Third, these agents may work to
“rewire” the death-inducing receptor/ligand pairs, which include Fas and FasL [23,
25, 26]. The second and third possibilities are not mutually exclusive and may work
in concert to result in tumor cell death (see Table 1).
Drug resistance is the leading cause of death in cancer [27]. Mechanisms that
have been suggested to account for drug resistance include overexpression of a
multidrug-resistance transporter (pgp -1) [27], failure to express death-inducing
receptors [27, 28], and our recent description of a metabolic strategy that may provide protection from a variety of stresses [24].
Our hypothesis is that mitochondrial metabolism in tumor cells significantly
influences susceptibility or resistance to Fas-dependent/anticancer agent–induced
tumor cell death. Our preliminary work provided the framework for the hypothesis that metabolic strategy confers resistance or susceptibility to apoptotic death of
tumor cells. We suggest that treatment of some cancer cells by chemotherapeutics
does indeed produce substantially enhanced costimulatory signals and that the treatment of cancer cells by chemotherapeutics and immune system involvement results
in substantially higher tumor cell death rates in the measured cases.
Figure 1 shows that treatment of L1210 tumor cells (a mouse leukemic cell
line [29]) cultured for 24 hours with methotrexate induces significant increases in
cell surface Fas. The lower concentration of 10–8 M is physiologically relevant to
chemotherapy because this is the upper limit of drug serum levels in humans. Our
results here are consistent with some of our previous published work where we
showed that a different costimulatory signal (B7.1) also increased in response to
methotrexate [2].
Figure 2 shows that different chemotherapeutic agents induce an increase in Fas,
but that different drugs can produce increases of varying amounts. This leads to the
tantalizing possibility that drug effectiveness could be correlated with levels of Fas
Table 1 Percent Death of L1210 cells with or without a 96-Hour Incubation with 10–8 M
Methotrexate and with or without an Exposure to Anti-Fas Antibodies∗
Percent death L1210 cells (%)
Anti-Fas exposure
No methotrexate
With methotrexate
No anti-Fas
Anti-Fas coated plates
1
7
4.72
79.98
∗
The results indicate that methotrexate sensitize leukemic cells to Fas-induced
death.
Uncoupling Cellular Respiration
151
Fig. 1 Expression of cell
surface Fas as a function of
treatment with a typical
oncolytic agent. There is a
significant increase in cell
surface Fas as a result of drug
treatment. The level of cell
surface Fas was determined
using fluorochromeconjugated anti-Fas
antibodies and flow
cytometry. The Fas levels are
measured relative to a control
for staining artifacts
expression and that this could be used to predict effective targeting of drugs via
Fas-induced destruction.
We have concentrated on Fas as an important example of the costimulatory signal; however, there are a number of other cell surface molecules that also play a role
in T-cell activation. In Fig. 3 , we show that B7.2 levels in HL60 (human leukemic
cells) also increase after treatment with 10–8 M Adriamycin. HL60 is a human cell
Fig. 2 Flow cytometry data showing increasing levels of cell surface Fas in L1210 cells as a result
of treatment with different chemotherapeutic agents. The horizontal axis shows intensity of the
fluorescence, and the vertical axis shows the number of cells at each intensity. Peaks that appear
farther to the right have higher levels of Fas per cell
152
M.K. Newell et al.
Fig. 3 Expression of the cell
surface costimulatory
molecule B7.2 as a function
of treatment with a typical
oncolytic agent. The level of
cell surface B7.2 was
determined using
fluorochrome-conjugated
anti-B7.2 antibodies and flow
cytometry. The B7.2 levels
are measured relative to a
control for staining artifacts
[31]
line, and the drug is different than that in previous figures. This implies that the
increase in the costimulatory signal as a result of drug treatment may be a general
phenomenon.
The data above strongly support the connection between the efficacy of drug
treatment and an increase in costimulatory immune signals. However, it is important
to know if this increase actually results in an increased immune-mediated death rate
for the tumor cells. In contrast with treatment with anti-Fas or methotrexate alone,
anti-Fas produces a striking increase in percent death for L1210 cells that have been
cultured in methotrexate. Therefore, drug treatments result in increased expression
of costimulatory signals and this, in turn, plays an important role in successful tumor
cell treatment.
Our second question concerned whether the treatment of tumor cells by
chemotherapeutics supplemented with agents that promote reactive oxygen intermediates result in increased susceptibility to immune-mediated cell death.
As we saw from the data above and from the work of others [28], chemotherapeutics cause an increase in the expression of costimulatory molecules. There are
other treatments that can also produce a similar increase. For example, Fig. 4 shows
the effects of culturing two tumor cell lines in subcytoxic concentrations of H2 O2 .
The H2 O2 treatment results in a threefold increase in the costimulatory molecule
B7.2. The increase from treatment with H2 O2 is also significantly larger that the
increase resulting from treatment with Adriamycin alone (Fig. 3).
Experiments with other cell types show that the expression of other costimulatory
molecules can also be substantially changed by the H2 O2 treatment. For example,
Fig. 5 shows that treatment of JB6 cells (a mouse fibroblast tumor cell line) with
H2 O2 causes up to a fourfold increase in cell surface Fas expression. Clearly what
Uncoupling Cellular Respiration
153
Fig. 4 Relative B7.2 levels
for two different tumor cell
lines with and without
addition of subcytotoxic
levels of H2 O2 . Obviously,
the addition of H2 O2
increases the expression of
cell surface costimulatory
molecules significantly
is needed is to look at the effectiveness of combined approaches, for example, using
both chemotherapeutics and H2 O2 .
Our third question dealt with whether drug resistance can be overcome by treatments that promote immune involvement in the destruction of tumor cells and
involved the hypothesis that uncoupling proteins regulate drug resistance by changing the metabolic state of a cell.
Fig. 5 Relative cell surface
Fas expression in JB6 cells as
a function of time after
exposure to subcytotoxic
levels of H2 O2 . We see,
again, that the levels of
costimulatory molecules
increase substantially after
exposure to H2 O2
154
M.K. Newell et al.
Fig. 6 A scatter diagram for drug-sensitive (left panel) and drug-resistant (right panel) tumor cells.
The horizontal axis is a measure of the change in pH across the inner mitochondrial membrane.
The vertical axis measures mitochondrial membrane potential. The results show that mitochondrial
membrane potential and pH difference is lower for the majority of drug-resistant cells. The data
are obtained from flow cytometry using a JC-1 dye
First we show that drug-sensitive cancer cells and drug-resistant tumor cells have
very different metabolic properties. As is well-known, cell metabolism is largely
governed by the action of the mitochondria. This, in turn, depends on some key
parameters that include the potential difference across the inner mitochondrial membrane and the change in acidity across the same membrane. Figure 6 shows a scatter
diagram for drug-sensitive and drug-resistant tumor cells indicating the range of
both of these parameters. The abscissa is a measure of the change in pH across
the inner mitochondrial membrane. The ordinate measures mitochondrial membrane potential. We see that the drug-resistant cells, in general, have a lower membrane potential, and the pH gradient is also significantly reduced for these cells.
These data have been substantiated and extended to other cell lines (including HL60
and HL60/MDR) by other measurements [24]. This provides strong evidence that
drug resistance is correlated with altered metabolic behavior.
A second example of how drug-resistant cells exhibit a different metabolic
behavior compared with drug-sensitive cells is provided by the rate of mitochondrial oleate consumption in these two different types of cells [24]. Figure 7 shows
that the drug-resistant cells typically have higher levels of oleate consumption, and
as glucose becomes limited, the rate of consumption increases dramatically in the
drug-resistant cells only. This argues that the ability to ”burn fat” is substantially
different between the two types of cells.
Uncoupling Cellular Respiration
155
Fig. 7 Rates of fatty acid (oleate) oxidation for drug-sensitive and drug-resistant cells. The oxidation rates for drug-resistant cells are generally higher than those for drug-sensitive cells, and this
becomes more pronounced at low glucose levels [24]
An important question is whether this metabolic difference leads to differences in
the way tumor cells present themselves to the immune system. In Fig. 8 , we present
evidence that drug-sensitive tumor cells (those in the left panels) typically display
relatively high levels of cell surface Fas, one of the requirements for Fas-induced
death [2, 30]. In contrast, the drug-resistant cells (seen in the right panels) show
virtually no cell surface Fas. These data suggest that the differences in metabolic
behavior are correlated with expression of cell surface Fas. Clearly, it is not known
if this is a cause-and-effect behavior or if it is mere correlation. Nonetheless, the
possibility of a causal relationship between cellular metabolism and the expression
of cell surface Fas may offer an intriguing tool for making tumor cells visible to the
immune system.
Treatment of drug-resistant tumor cells with chemotherapeutic agents does not
increase expression of cell surface Fas or other costimulatory molecules. This
behavior is illustrated in Fig. 9 below. This is in stark contrast with the behavior
of drug-sensitive tumor cells (see Fig. 2, Fig. 3, and Fig. 4), which do show substantial increases in Fas with chemotherapeutic treatments.
The results above suggest that one possible reason that chemotherapy is ineffective for drug-resistant cells is that the costimulatory signal is absent. This naturally leads to the question of whether costimulation, particularly in drug-resistant
cells, could be increased by other treatments. In Fig. 10 , we show a summary of
experiments on three different (relatively) drug-resistant human cell lines. We see
156
M.K. Newell et al.
Fig. 8 Levels of cell surface Fas expression on drug-sensitive (left panels) and drug-resistant
(right panels) cells as measured by flow cytometry. The horizontal axis shows intensity of the
fluorescence, and the vertical axis shows the number of cells at each level of fluorescence intensity.
Peaks that appear farther to the right have higher levels of Fas per cell. Note that the drug-sensitive
cells have significantly higher levels of cell surface Fas
that treatment with subcytotoxic H2 O2 increases B7.2 expression in all cases and
that it is increased by a factor of 2 to 3 in some cases. This increase is, in fact, very
similar to that for the drug-sensitive cells (see Fig. 4 for example).
We have suggested that differences in metabolic behavior correlate with differences in sensitivity to drugs [24]. One possible mechanism that would account for
this could be the activity of the uncoupling protein UCP2. These proteins have
been shown to produce a number of metabolic changes that are consistent with
the metabolic behavior of drug-resistant cells. We measured the presence of UCP2
in mitochondria of drug-resistant and drug-sensitive cells by isolating mitochondria and using antibodies to UCP2 in Western blot analysis (Fig. 11 ). Significantly
higher levels of UCP2 were detected consistently in the resistant cell lines (lanes 4
and 5). These results provide the rationale for an investigation of the role of uncoupling proteins as a part of the mechanism for drug resistance. If this mechanism
were confirmed, this would create a new approach for treatments of drug-resistant
cancers.
Uncoupling Cellular Respiration
Fig. 9 Levels of cell surface
Fas expression on
drug-sensitive L1210 (upper
panel) and drug-resistant
L1210/DDP (lower panel)
cells as measured by flow
cytometry. Cells were
cultured in the presence of
chemotherapeutic agents
overnight. The drug-sensitive
cells show increases in Fas as
a result of the drug. In
contrast, the drug-resistant
cells are unchanged with
treatment. The horizontal axis
shows intensity of the
fluorescence, and the vertical
axis shows the number of
cells at each intensity. Peaks
that appear farther to the right
have higher levels of Fas per
cell. Note that the
drug-sensitive cells have
significantly higher levels of
cell surface Fas
Fig. 10 Relative expression of
the costimulatory molecule B7.2
on three different drug-resistant
human cell lines with and without
treatment with subcytotoxic
amounts of H2 O2 . In contrast
with treatments with
chemotherapeutics alone, we see
here a substantial increase in the
costimulatory signal, which is
achieved by adding the H2 O2
157
158
M.K. Newell et al.
Fig. 11 Western blot analysis for
UCP2 of mitochondrial extracts
from drug-sensitive L1210 (lanes
2 and 3) and drug-resistant
L1210/DDP (lanes 4 and 5) tumor
cells. UCP2 levels are
substantially larger in the
drug-resistant cells [24]
9 Therapeutic Possibilities
As an example of using a metabolic approach to provide therapeutic benefit, we will
discuss the performance of one compound, dichloroacetate, in detail. There are two
overall goals: first, to slow or inhibit high-rate glycolysis, and the second, to block
fatty acid oxidation at several key enzymatic steps.
We and others have identified 2-deoxyglucose (2DG) as a key regulator of highrate glycolysis. However, our work has demonstrated that for most rapidly dividing
tumor cells, treatment with 2DG alone does not cause tumor cell death. In fact, in
terms of percent viability, treatment with 2DG results in improved percent viability
for the vast majority of tumors. The mechanism appears to be that the compound
results in growth arrest of the cells. In combination with traditional chemotherapeutic agents, the addition of 2DG increases the oncolytic activity of the therapeutic.
Though beneficial, this approach leaves open the likely possibility that treatment
with 2DG and chemotherapeutic agents such as Taxol (paclitaxel) or Adriamycin
will eventually result in drug resistance, the leading cause of death due to cancer.
Targeting metabolism as a chemotherapeutic approach is a major focus of our
work, and we focus on the inhibition of tumor cell–specific fatty acid oxidation.
Toward that end, we have used the several fatty acid oxidation (FAO) inhibitors.
We have extensive data showing the effectiveness of this approach in inducing the
apoptotic death of a wide variety of tumor types.
We note the listing of dichloroacetate (DCA) in each of these two categories of
metabolic modifications: as a glycolytic inhibitor as well as a fatty acid oxidation
inhibitor. With metabolic strategies, one must consider that cells have feedback strategies that communicate when energy demands are met or when energy demands are
unmet. The cell will respond accordingly with either catabolic or anabolic pathways
to provide optimization of energy expenditure. Because DCA is similar in structure
to pyruvate, the molecule can act as a negative regulator of glycolytic processes.
In addition, the structure of DCA is known to fit the negative regulatory binding
pocket of pyruvate dehydrogenase kinase, an enzyme that increases under conditions of starvation promoting fatty acid oxidation when glucose or carbohydrate
Uncoupling Cellular Respiration
159
stores are limited. Thus, DCA will act as a negative regulator of fatty acid oxidation, a key strategy of drug-resistant tumor cells. This selective interference causes
cells that are normally selectively nonapoptotic to now die from apoptotic cell death.
While theoretically promising, the effects of DCA when used in combination with current “standard of care” chemotherapeutics need to be rigorously tested for potentially
harmful or toxic side effects.
In summary, the technique of metabolic disruption shows promise for creating
therapeutic strategies that may help combat cancer or other diseases. These strategies have the potential of changing the way the immune system interacts with tumors
and other metabolically altered tissues.
References
1. Voet D, Voet J. Chapter 22: Electron Transport and Oxidative Phosphorylation. Biochemistry
John Wiley & Sons, Inc. (2004).
2. Bhushan A, et al. Drug resistance results in alterations in expression of immune recognition
molecules and failure to express Fas (CD95). Immunol Cell Biol1998; 76:350–356.
3. Brundtland GH. The Global Burden of Cancer In: Stewart BW, Kleihues P, eds. World cancer report. Lyon: International Agency for Research on Cancer, World Health Organization,
2003:352.
4. Marrack P, Kappler J. The T cell receptor. Science 1987; 238:1073–1079.
5. Bretscher PA, Cohn M. A theory of self-nonself discrimination. Science 1970; 169:
1042–1049.
6. Linsley PS, Ledbetter JA. The role of the CD28 receptor during T cell responses to antigen.
Ann Rev Immunol 1993; 11:191–212.
7. Linsley PS, et al. Human B7-1 (CD80) and B7-2 (CD86) bind with similar avidities but distinct kinetics to CD28 and CTLA4. Immunity 1994; 1:793–801.
8. June CH, et al. The B7 and CD28 receptor families. Immunol Today 1994; 15:321–330.
9. Kuchroo VK, et al. B7-1 and B7-2 costimulatory molecules activate differentially the
Th1/Th2 developmental pathways: application to autoimmune disease therapy. Cell 1995; 80:
707–718.
10. Lanier LL, et al. CD80(B7) and CD86(B70) provide similiar costimulatory signals for
T cell proliferation, cytokine production, and generation of CTL. J Immunol 1995; 154:
97–105.
11. Alderson MR, et al. Fas transduces activation signals in normal human T lymphocytes. J Exp
Med 1993; 178:2231–2235.
12. Nagata S, Human autoimmune lymphoproliferative syndrome, a defect in the apoptosisinducing Fas receptor: a lesson from the mouse model. J Hum Genet 1998; 43(1):2–8.
13. Desbarats J, et al. Dichotomy between na¨ıve and memory CD4+ T cell responses to Fas
(CD95) engagement. Proc Natl Acad Sci USA 1999; 96:8104–8109.
14. Desbarats J, Newell MK. Fas engagement accelerates liver regeneration after partial hepatectomy. Nat Med 2000; 6(8):920–923.
15. Mavligit GM, et al. Cell-mediated immunity to human solid tumors: in vitro detection by lymphocyte blastogenic responses to cell-associated and solubilized tumor antigens. Natl Cancer
Inst Monogr 1973; 37:167–176.
16. Whelan M, et al. Cancer immunotherapy: an embarrassment of riches? Drug Discov Today
2003; 8(6):253–258.
17. Martindale D. T cell triumph: immunotherapy may have finally turned a corner. Sci Am 2003;
288(2):18–19.
160
M.K. Newell et al.
18. Tsuruo T, et al. Molecular targeting therapy of cancer: drug resistance, apoptosis and survival
signal. Cancer Sci 2003; 94(1):15–21.
19. Newell MK, et al. Does the oxidative/glycolytic ratio determine proliferation or death in
immune recognition? Ann NY Acad Sci 1999; 887:77–82.
20. Nagata S, Golstein P. The Fas death factor. Science 1995; 267:1449–1456.
21. Nagata S. Apoptosis by death factor. Cell 1997; 88:355–365.
22. Schneider P, Tschopp J. Apoptosis induced by death receptors. Pharm Acta Helv 2000;
74(2–3):281–286.
23. Kataoka T, et al. Expression level of c-FLIP versus Fas determines susceptibility to Fas ligandinduced cell death in murine thymoma EL-4 cells. Exp Cell Res 2002; 273(2):256–264.
24. Harper M-E, et al. Characterization of a novel metabolic strategy used by drug-resistant tumor
cells. FASEB J 2002;16(12):1550–1557.
25. Sinkovics JG, Horvath JC. Virological and immunological connotations of apoptotic and antiapoptotic forces in neoplasia. Int J Oncol 2001; 19(3):473–488.
26. Green DR, Evan GI. A matter of life and death. Cancer Cell 2002; 1(1):19–30.
27. Tolomeo M, Simoni D. Drug resistance and apoptosis in cancer treatment: development of
new apoptosis-inducing agents active in drug resistant malignancies. Curr Med Chem AntiCancer Agents 2002; 2(3):387–401.
28. Landowski TH, et al. Myeloma cells selected for resistance to CD95-mediated apoptosis are
not cross-resistant to cytotoxic drugs: evidence for independent mechanisms of caspase activation. Blood 1999; 94(1):265–274.
29. Venditti JM, Sheldon DR, Goldin A. Evaluation of antileukemic agents employing advanced
leukemia L1210 in mice. VII. Cancer Res 1964; 24(1 Pt 1):145–210.
30. Villalobos-Menuey EM. Metabolic regulation of the cellular distribution and function of Fas
(CD95). Master’s thesis, University of Colorado, Colorado Springs, 2001:1–83.
31. Newell MK, Melamede RJ, Villalobos-Menuey E, Swartzendruber D, Trauger R, Camley RE,
Crisp W. The effects of chemotherapeutics on cellular metabolism and consequent immune
recognition. J Immune Based Ther Vaccines 2004; 2(1):3.
How Cancer Cells Escape Death
Erica Werner
Abstract Programmed cell death, or apoptosis, is a cell-autonomous mechanism
to restrict deviation from normal cellular function. Central to this mechanism are
Bcl-2 proteins, which monitor normal cellular function to define apoptotic fate
by surveying, integrating, and decoding multiple survival and proapoptotic signals
from the intra and extra cellular environment. Escape from these restrictive mechanisms is essential to accommodate the cellular changes driving tumor establishment
and development. To do so, many of the factors promoting tumorigenesis, such as
genetic lesions, hypoxia, p53 mutations and oncogene activation, gain control over
the main apoptotic pathway to promote cell survival. For this reason, the apoptotic
pathway is the focus of remarkable interest for targeted cancer therapy development.
However, as apoptosis evasion is a tumor promoting factor itself, the results of first
therapeutic targeting attempts reveal that these targets pose drawbacks similar to
the exhibited by classic oncogenic targets, including the rapid development of resistance through high mutation rate, intra and inter tumor heterogeneity and narrow
effective range.
In addition to tumor promoting pathways, apoptosis evasion is fostered by factors
secondary to transformation and tumor development. These factors, such as changes
in the interaction with the extracellular environment and metabolism, have received
less attention because of their underestimated role in transformation but might prove
to be fruitful avenues for intervention as they are common to many tissues and affect
apoptotic fate under the influence of reduced oncogenic pressure. The significance
of these factors with respect to apoptosis evasion and their potential impact for therapy development are discussed.
Keywords
Glycolysis
Apoptosis · Bcl-2 · Mitochondria · Prosurvival · Environment ·
E. Werner (B)
Department of Cell Biology, School of Medicine, Emory University, Atlanta, GA 30322
e-mail: ericaw@cellbio.emory.edu
S.P. Apte, R. Sarangarajan (eds.), Cellular Respiration and Carcinogenesis,
DOI 10.1007/978-1-59745-435-3 12,
C Humana Press, a part of Springer Science+Business Media, LLC 2009
161
162
E. Werner
1 Introduction
Death is universal and certain. At the cellular level, this fate is controlled by apoptosis, a genetic program present in every cell type of organisms, including cancer cells.
Evasion of this program is deemed key to carcinogenesis and tumor progression.
Thus, extraordinary efforts have been committed to understand how cancer cells
evade the apoptotic pathway and thereby influence disease prognosis and targeted
therapy design. Although past and current studies have consistently revealed the significant role of apoptosis evasion in the process of carcinogenesis, correlations of
disease progression and prognosis with the expression of apoptosis-related markers
have been inconsistent, and therapies targeting effectors of the apoptotic pathway
have been of limited efficacy [1]. This suggests that additional cell autonomous or
extracellular mechanisms are likely contributing to apoptotic outcome. This will be
the major focus in this chapter.
The apoptotic pathway is a surveillance mechanism monitoring cell activity and
imposing strict limits to permissible cellular function. These limits are constantly
challenged during cancer initiation and progression. Early deregulated proliferation demands multifaceted changes encompassing intracellular signaling pathways,
metabolism, nutrient and energy requirements, and tissue architecture, which trigger
intracellular and extracellular proapoptotic signals in normal cells. At later stages
of tumor development, cancer expansion and metastasis adds a further cohort of
apoptosis-inducing signals, generated in opposition to changes in the function of
multiple cell types and altered environments.
In contrast with mutations in oncogene and tumor suppressor genes, evasion of
the apoptotic pathway does not initiate and drive carcinogenesis per se, but is instrumental for tumor cells to successfully thrive despite harboring alterations that would
not be tolerated in normal circumstances. The mechanisms driving tumor progression are integrated with apoptosis resistance acquisition, at least at three levels, so
as to ensure successful apoptosis evasion:
Somatic mutations, promoter methylation, chromosomal translocations and deletions associated with the inherent genetic instability of cancer cells alter the expression of oncogenes as well as those of key apoptotic players. Even though this
has been the most explored mechanism to explain the increased apoptosis resistance of tumor cells [1], correlations in most cases are difficult to generalize and
in many cases are nonconfirmatory. One interpretation of this outcome could be
that activity rather than expression levels are consequential for carcinogenesis, as
the apoptotic players are mainly regulated through posttranslational mechanisms,
such as phosphorylation or ubiquitin-targeted proteasome-dependent degradation.
Another reason may be associated with the difficulty to evaluate whether the identified changes in expression are indeed a causal factor in disease progression or are
merely coincidental.
A second factor promoting apoptosis resistance is an overall shift in the nature
of signals surveyed by the apoptotic machinery, reinforcing prosurvival signals
while interfering with proapoptotic signals (see Fig. 1). This altered input is the
direct result of two concurrent processes; one leading to independence from an
How Cancer Cells Escape Death
163
Fig. 1 Apoptosis input is balanced by prosurvival (gray) and proapoptotic (black) signals. These
signals control the activity (crossed arrows represent inhibitory signals, and pointed arrows represent activation; dashed lines represent loss of that particular connection) of antiapoptotic Bcl-2
members (round boxes) and proapoptotic Bcl-2 family members (square boxes). (A) In normal
cells, the extracellular environment through growth factors, cell adhesion, and cytoskeleton organization restrains activity of proapoptotic Bcl-2 family members. P53 responds to endogenous
stress providing proapoptotic signals. (B) During carcinogenesis, control of the apoptotic pathway is taken over by oncogene-activated kinases, prosurvival cell-surface interactions, glycolysis
activity, and loss of p53 function
164
E. Werner
environment rich in proapoptotic signals and the other altering the intracellular signal transduction network through oncogene activity and tumor suppressor silencing.
The marked influence of these factors on apoptotic fate becomes apparent as apoptosis sensitivity can be restored by normalizing cancer cell communication with the
extracellular environment or by neutralizing oncogene activity (see Section 3).
A third, far less studied factor affecting apoptotic fate is the characteristic altered
metabolism associated with transformed and tumor cells. Mitochondria are a central piece in the apoptotic pathway, but are a control center for normal and tumor
metabolism as well. The importance of metabolism in carcinogenesis is gaining
recognition and promises a source of potentially viable therapeutic approaches, as
many of the involved pathways control enzymatic activities suitable for pharmacologic inhibition are linked to apoptosis.
Because of space constrains, in this chapter I will only outline the essential players and concepts necessary to provide a conceptual framework illustrated by a few
examples and will cite only select key reviews to refer the reader to further details
in each section.
2 The Bare Pathway to Death
The essential apoptotic pathway responds to specific triggers, such as proapoptotic
extracellular ligands, cellular stressors, or deprivation of a survival signal. These
elicitors activate an intracellular proteases cascade in charge of cleaving multiple
targets to cause cell death. The resulting cell debris, organelle and nucleus fragments
are packed into apoptotic bodies to be cleanly disposed by professional phagocytes
and/or neighboring cells thereby averting inflammatory responses [2].
At the center of the apoptotic pathway is a regulatory step dictated by the proteinprotein interaction of the Bcl-2 family members, controlled by proapoptotic and
prosurvival signals generated in the extracellular environment (extrinsic pathway)
and inside the cell (intrinsic pathway). If the activity balance of Bcl-2 proteins is
inclined toward cell death, proapoptotic members redistribute to mitochondria and
cause outer membrane permeability, with the consequent release of proteins contained in the mitochondrial intermembrane space [3]. The proteins released into the
cytosol activate caspases and other regulators/effectors of cell breakdown.
2.1 The Family of Bcl-2 Proteins: Initiators of the Intrinsic
Pathway
Cellular stress signals such as growth factor deprivation, DNA damage, oncogene
activation, or hypoxia trigger apoptosis by regulating the activity of members of the
Bcl-2 protein family. Each member shares a protein-protein interaction domain with
homology to Bcl-2 (BH domains) and is grouped according to its role in apoptosis:
multidomain antiapoptotic (Bcl-2, Bcl-xL, Bcl-w, Mcl-1, and Bfl-1/A1), multidomain proapoptotic (Bax and Bak), and BH3-only proapoptotic (Bid, Bim, Bad, Bik,
How Cancer Cells Escape Death
165
Noxa, Puma, Bmf, and HRK). Tissue-specific expression and posttranslational regulation contribute to establish an activity balance of these proteins, which mediates
stimuli-selective and tissue-specific induction of apoptosis. The signals that influence apoptosis converge to control the activity of proapoptotic Bax and Bak, as
genetic deletion of both proteins renders cells unable to undergo apoptosis via the
intrinsic pathway [4]. When the balance favors activity of proapoptotic members,
Bax translocates from the cytosol to the mitochondrial outer membrane, where it
polymerizes with mitochondria-localized Bax to form a pore. Bcl-2 family members modulate this step by two alternative but nonexclusive mechanisms. Some signals induce apoptosis by activating proapoptotic BH3-only proteins, which bind to
and neutralize selectively antiapoptotic Bcl-2 family members, releasing Bak/Bax to
oligomerize. For example, proapoptotic Bad mediates induction of apoptosis by this
mechanism [5]. Other proapoptotic BH3-only proteins such as Bid and Bim compete with antiapoptotic members but also have the potential to interact directly with
Bax/Bak for activation. In both scenarios, the emerging consensus is that several
of the antiapoptotic factors need to be simultaneously neutralized by proapoptotic
factors for apoptosis to proceed, requiring the cooperation of multiple members
of the family acting as sensitizers or de-repressors. Thus the current view is that
because proapoptotic BH3-only members present a selective pattern of interaction
with antiapoptotic Bcl-2 family members, the right combination of antiapoptotic and
proapoptotic members has to be engaged to either halt or unleash apoptosis. These
interactions between Bcl-2 family members are further regulated by phosphorylation status and/or targeted degradation (see Fig. 1 and below and Ref. 3).
2.2 Mitochondrial Outer Membrane Permeability
Once Bak/Bax become activated, they increase mitochondrial outer membrane permeability to release proteins contained in the intermembrane space. This event
causes mitochondrial membrane potential dissipation and matrix swelling, compromising mitochondrial function and structure. This is the mechanistically least understood step in the pathway, as the exact identity of the pore releasing the proteins is
still a matter of debate [6]. A candidate participant and modulator is the permeability
transition pore (PTP), a multiprotein complex spanning the outer and inner mitochondrial membrane formed by voltage-dependent anion channel (VDAC), adenine nucleotide translocase, benzodiazepine receptor, cyclophilin D, hexokinase,
and other proteins. This pore mediates the exchange of ATP/ADP and other solutes
in the normal state. Its participation in apoptosis is supported by the effects of pharmacologic inhibitors and activators for the different components and thus constitutes
an attractive target for therapy bypassing Bcl-2–dependent regulation [7]. However,
these inhibitors do not abolish all pathways conducive to apoptosis, and apoptosis still occurs in models with genetic deficiencies of these proteins, suggesting that
this complex regulates other pores releasing the apoptotic effectors or that—in addition to the PTP—there are multiple mechanisms for proapoptotic effectors release.
Regardless of the mechanism, however, once the mitochondrial membrane becomes
166
E. Werner
permeable, the cell is committed to die even if downstream caspase activity is abrogated with pharmacologic inhibitors [8].
Proteins residing in the intermembrane space such as cytochrome c, Smac/
DIABLO, apoptosis-inducing factor (AIF), Endo G, and Omi/HtrA2 are released
simultaneously and shortly after Bax/Bak translocation. Free in the cytoplasm,
cytochrome c binds Apaf-1, causing a conformational change facilitating ATP/dATP
exchange and oligomerization followed by recruitment of caspase 9. This multiprotein complex, the apoptosome, is formed to activate the initiator caspase 9, which
cleaves and activates the downstream effector caspases 3 and 7. Smac/DIABLO
and Omi/HtrA2 are proteases that inactivate the endogenous cytosolic inhibitors
IAPs (inhibitors of apoptosis) to facilitate the activity of the newly formed apoptosome [9].
AIF as well as Endo G translocate to the nucleus and participate in chromatin
condensation and lysis [2]. The extent of their contribution to the decision to
undergo apoptosis is still unclear, but they are significant contributors to biochemical and morphologic manifestations of apoptosis. AIF is a flavoprotein participating
in normal mitochondrial bioenergetic and redox metabolism, thus, its release contributes to mitochondrial dysfunction and to apoptosis irreversibility [10].
2.3 Caspases Are the Executors of Apoptosis
Active cell dismantling is brought about by proteases termed caspases, which use a
nucleophilic cysteine in the active site to cleave immediately distal to an aspartatecontaining motif present in specific substrates. Among more than 14 members of
this family in mammals, only 7 play known roles in apoptosis, while many participate in physiologic functions unrelated to apoptosis. All caspases exist as zymogens
in the cytosol and are grouped according to their mechanism of activation. The initiator or apical caspases (caspase 2, 8, 9, and 10) are autocatalytically activated by
proximity after associating through their N-terminal domain to a multiprotein activating complex incrementing their catalytic activity by several orders of magnitude.
Effector caspases (3, 6, and 7) are activated by initiator caspases–mediated cleavage
in a 20- to 30-amino-acid linker sequence [11].
A further level of regulation comes from the activity of cytosolic IAPs, which
neutralize caspase activity. Discovered for their ability to suppress apoptosis
in baculovirus-infected cells, the IAP proteins are regulated by ubiquitylationmediated proteasome degradation [12].
2.4 Changes in Apoptotic Players Associated with Cancer
Unlike oncogenic mutations, changes in the apoptotic pathway are generally insufficient to initiate the carcinogenesis process but play a significant role in tumor
promotion. Overexpression of antiapoptotic Bcl-2 or Bcl-xL is observed in multiple
cancers and accounts for acquired chemoresistance [3]. Bcl-2 was first identified
as an oncogene at a chromosomal breakpoint of t(14:18) in human follicular B-cell
lymphoma. Unexpectedly, this oncogene localized to the mitochondrion and did not
How Cancer Cells Escape Death
167
induce cell proliferation. Accordingly, when Bcl-2 is overexpressed as a transgene
in murine models of carcinogenesis, it is not sufficient to elicit oncogenesis but promotes clonogenic survival and growth of the primary tumor when expressed together
with an archetypal oncogene. Bcl-2 coexpression with Myc drives tumorigenesis in
lymphocytes and in mammary gland. Further studies conditioning Bcl-2 expression
in a double transgenic mouse together with Myc shows that switching Bcl-2 expression off does not alter the onset of Myc-induced leukemia but results in tumor remission and prolonged mice survival. These experiments highlight the tumor-promoting
function of antiapoptotic factors such as Bcl-2 while simultaneously revealing the
persistence and continuous pressure of proapoptotic signals acting on a tumor.
Contrary to expectations, deletion of apoptosome components does not contribute to oncogenic transformation. This is evidenced by experiments where the
selective deficiency of Apaf-1 or caspase 9 does not enhance lymphomagenesis nor
contributes to fibroblast transformation [13]. Analysis of tumor samples reveal that
alterations in molecules regulating the apoptosome function are more frequent [9].
From these studies stands out a correlation between the increased expression in
tumors of caspases inhibitors (IAPs), especially survivin, with poor disease prognosis. Enhanced survival capabilities become significant at advanced stages of tumor
progression during invasion and metastasis. In fact, isolated prostate tumor cells circulating in the bloodstream in animals bearing an orthotopic xenograft of prostate
carcinoma cells showed increased resistance to anoikis, a special type of cell death
(see Section 3.1) due to induction of the caspase inhibitors XIAP, cIAP2, and survivin, correlating with enhanced metastatic potential [14]. Survivin is a member of
the AIF protein family that interferes with caspase activity and is one of the most
commonly expressed proteins in transformed cell lines and human tumors, while
absent in normal tissues [15]. Moreover, it is a poor prognosis factor correlating
with aggressiveness and invasive behavior. Survivin is a target of p53 transcriptional
repression (see Section 4.1), therefore it is found upregulated in tumors bearing p53
mutations. Targeting survivin for therapeutic purposes, however, is problematic due
to the essential role it plays in mitosis.
3 Deceiving the Surveillance Mechanism by Upregulating
Survival Signals
The activity of Bcl-2 family members is regulated by information collected from the
intracellular and extracellular milieu, both of which undergo significant alterations
during carcinogenesis.
3.1 Evading Restrictions Imposed by Cell-Cell Interactions
and Tissue Architecture
Multicellular organisms have a number of signaling mechanisms geared toward
ensuring systematic cell proliferation and differentiation during development and
limited clonal expansion of somatic cells during growth and cell renewal in the
168
E. Werner
adult organism. Furthermore, proper tissue organization is maintained by the spatially restricted availability of proliferation and prosurvival signals from the basement membrane and by repressive proapoptotic signals from neighboring cells.
Repressive signals from adjacent cells actively promote apoptosis directing the
selective death of cell populations to organize organs and tissues during development, as is the case for eliminating abnormal, misplaced, or harmful cells during
lymphocyte T and B development or for neuronal pruning during development. In
the adult state, controlled apoptosis actively maintains tissue structures. For example, in vivo and in vitro experiments show that apoptosis actively maintains the
lumen in the mammary gland acini through a signal that promotes proapoptotic
Bim activity. A lumen is maintained even when disorganization is experimentally
introduced by forced expression of proliferation inducing proteins leading to abnormal cell multiplication in the acini. Lost of structure by lumen repopulation can
occur only when survival signals are delivered either by the coexpression of antiapoptotic proteins or by an oncogene conferring both prosurvival and proliferative
signals [16]. Thus, these experiments illustrate the existence of geometric cues built
in the organization of tissues, which need to be surmounted very early in tumor
development.
Cell surface molecules involved in adhesion communicate these controlling signals, acting upon later stages in tumor development as well. In multistage cancer
models such as chemically induced squamous cell carcinoma, expression of survivin as a transgene in skin opposes apoptotic signaling from surrounding cells,
thereby significantly reducing the spontaneous regression rate of benign lesions
[17]. The intercellular adhesion molecule E-cadherin mediates tissue architecture
restrictive signals that are active during cancer initiation as well as during metastasis,
as downregulation of E-cadherin expression correlates with skin carcinoma progression. Accordingly, restitution of E-cadherin expression in transformed keratinocytes
reduces in vivo tumor formation potential [18]. E-cadherin silencing in mammary
epithelial cells, accelerates invasive disease and metastasis formation in transgenic
mice models, forming tumors resembling invasive lobular carcinoma [19].
An additional source of restrictive signals is the obligatory epithelial cell contact with the basement membrane, which delivers localized proliferation and prosurvival signals, actively opposing apoptosis [20]. When epithelial cells detach and
lose contact, they undergo a special form of cell death designated anoikis. This
mechanism maintains digestive epithelia homeostasis by inducing death and shedding of enterocytes at the intestinal villi tip. In contrast, it is not observed in highly
motile cells such as fibroblasts, where interference with integrin-matrix interactions
triggers instead a tissue repair and remodeling phenotype characterized by the activation of NF-␬B and matrix metalloproteinase release [21].
Cell detachment causes integrin disengagement from the extracellular matrix,
focal adhesion complex disassembly, and cytoskeleton disorganization, triggering
multiple proapoptotic signals. Interference with ␤1 integrin binding to the extracellular matrix using function blocking antibodies is sufficient to induce keratinocyte
apoptosis characterized by caspase 8 activation and proapoptotic Bid translocation to mitochondria [22]. Disassembly of focal adhesion complexes disrupts
How Cancer Cells Escape Death
169
prosurvival signaling mediated by FAK, SRC, and PI3K-AKT kinases, all recruited
and activated by organized focal adhesion complexes in substrate-attached cells.
Interference with FAK activity induces rapid Bax translocation to mitochondria
in mammary gland epithelial cells [23]. The disorganization of microfilament and
microtubule cytoskeleton during cell detachment also releases proapoptotic BH3only proteins Bmf and Bim that neutralize Bcl-2 antiapoptotic activity. Proapoptotic
Bmf is indirectly associated to the actin cytoskeleton by binding to dynein light
chain 2 complexed to myosin V. Proapoptotic Bim associates indirectly to tubulin
through dynein light chain 1 sequestered by associated dynein motor complex (see
Ref. 24 and references therein). Resistance to anoikis is a key factor to cell survival
during invasion, metastasis, and growth in new environments at distal places from
the primary tumor. Consequently with their prominent role in anoikis prevention,
suppression of FAK activity diminishes resistance to anoikis as well as metastatic
potential [25].
An additional strategy used by tumor cells to escape apoptosis is to alter the cell
surface repertoire of receptors. The apoptotic balance is modulated either by silencing cell surface molecules mediating proapoptotic signals in non-transformed cells
and/or by inducing new molecules promoting cell survival. The receptor for netrin1, NHC5 or “deleted in colon cancer,” induces apoptosis when unengaged, thus its
expression is reduced in most colorectal cancer types and many other tumor types
[26]. Because survival signaling is integrin-selective, squamous cell skin carcinomas
switch the expression of integrin subunit, repressing the anoikis-promoting alphav
beta5 and inducing alphav beta6, which renders these cells protected by increasing
AKT activation [27]. However, in leukemias the significance of cell adhesion and
the environment in the control of apoptosis is indisputably most evident: These normally nonadherent cells exhibit a much higher sensitivity to apoptosis when tested
separated from their environment in vitro than when tested in vivo [28]. In fact,
␤1 integrin–dependent adhesion of leukemia cells in vitro is sufficient to induce
proapoptotic Bim destabilization and confer resistance to apoptosis. A role for the
environment is further supported by experiments showing that multiple myeloma
cells become significantly more resistant to chemotherapeutic drugs when adhering
to bone marrow stroma [29].
3.2 Autonomous Survival Signals Acquired Through Oncogene
Activation and Anti-oncogene Silencing
The remarkable effects of oncogene-targeted therapeutic agents used in mouse
tumor models reveal the dominant influence of an activated oncogene on tumor
phenotype and survival. This has led to the concept of “oncogene addiction” [30].
Most oncogene-activated and tumor suppressor–silenced pathways drive proliferation and promote cell survival by increasing the activity of ERK, PI3K, and AKT
kinases [31]. The activation of these kinases downstream of Src, Abl, and EGFR
increases prosurvival signaling [30]. Mutations in the effectors of these pathways
either increase their catalytic activity or silence the activity of negative regulators.
170
E. Werner
These kinases have a profound effect on the function of the apoptotic cascade by
regulating the expression, function, and stability of Bcl2 family members through
transcriptional and posttranslational mechanisms. ERK induces the expression of
antiapoptotic proteins Bcl-2, Bcl-xL, and Mcl-1, while MAPK-dependent phosphorylation stabilizes Bcl-2 and causes proapoptotic Bim degradation [32].
AKT is the active kinase most frequently found in tumors and mediates prosurvival signals from many oncogenes and growth factors [33]. AKT exerts a dominant
prosurvival activity through multiple mechanisms. AKT phosphorylates the BH3
domain of proapoptotic Bad, interfering with apoptotic function and promoting
sequestration by cytosolic 14-3-3 proteins [34]. In addition, AKT regulates glucose
metabolism thereby amplifying prosurvival signals (see Section 4.2).
4 Survival Effects of Metabolic Changes in Cancer Cells
Unrestricted proliferation amid accumulated mutations drive the accrual of profound alterations in cell metabolism and function, in addition to signal transduction.
Furthermore, metabolic changes are advanced by stress that follows local hypoxia
and metabolite accumulation imposed by the substantial and rapid increment in cell
mass. There is mounting evidence to suggest that these abnormalities, which would
be a source of stress and proapoptotic signals in normal cells, are not only opposed
by tumors but also are key in fostering the acquisition of survival capabilities and
are a determinant factor in tumor promotion and expansion [35].
4.1 p53
When normal cells experience alterations in the cell cycle and metabolism, such as
reactive oxygen species, hypoxia, double-strand DNA breaks, and abnormal proliferation rate, p53 tumor suppressor is activated to induce growth arrest and/or apoptosis to ensure DNA repair in order to proceed with cell replication. P53 induces
apoptosis by multiple mechanisms. It promotes apoptosis through the transcriptional induction of the proapoptotic BH3-only proteins Bad, Bax, Puma, and Noxa.
In a second mechanism, p53 activates Bax directly, releasing proapoptotic BH3only proteins sequestered by Bcl-X and promoting Bak oligomerization by displacing bound antiapoptotic Mcl-1. P53 activation can be detected in pre-neoplastic
lesions, including colon adenomas, breast carcinomas in situ, and lung hyperplasias in response to replicative stress resulting from deregulated cell proliferation
and to double-strand DNA breaks, but preceding any signs of genomic instability or genetic alterations in the p53 pathway [36]. This observation indicates that
p53 is activated very early in carcinogenesis by oncogenic stress and acts as an
apoptosis-promoter and a barrier for tumor development. This barrier, however,
is early averted by inactivating or attenuating p53, as mutations can be detected
in early dysplastic lesions of non–small cell lung carcinomas as well as in most
human cancers, thereby driving tumor development [37]. Experiments in models
How Cancer Cells Escape Death
171
of myc-driven lymphomas show that the apoptosis promoting activity is sufficient
for effective p53-dependent tumor suppression, while growth arrest function is dispensable [38]. Restitution of p53 expression in advanced tumors causes significant
tumor regression by inducing apoptosis in lymphomas and senescence in sarcomas and liver carcinomas [39]. Thus, these experiments underscore the significant
role of p53 in suppressing tumor development through the control of the apoptotic
pathway.
4.2 Glycolysis
A switch to glycolysis as an alternative or additional source for ATP is essential
to the cancer phenotype. In normal cells, p53 imposes a limitation on the glycolytic rate by a two-pronged mechanism [40]. P53 induces the expression of a
cytochrome oxidase regulatory subunit, synthesis of cytochrome C oxidase, increasing mitochondrial respiration, and induces the expression of Tigar, which controls
fructose-2,6-biphosphate intracellular levels, inhibiting glycolysis at the level of
6-phosphofructo-1-kinase activity. During transformation, glycolysis increases by
several mechanisms in addition to loss of p53 control. The levels of fructose-2,6biphosphate are regulated by several kinases, including AKT, which phosphorylates
and activates 6-phosphofructo-2 kinase to increase glycolysis. Local hypoxic conditions, as well as alterations in metabolism resulting from accrued mutations and
from oncogene activation, such as Bcr-Abl, Myc, and K-Ras, can trigger a switch to
glycolysis [41].
Glucose utilization is a survival factor per se, conferring resistance to apoptosis
proportional to the glycolytic rate [42]. Glycolysis inhibition in Myc-transformed
cells precipitates apoptosis, whereas normal cells respond with cell cycle arrest [43].
Although the mechanism is not entirely clear yet, and glycolysis could modulate
several steps in the apoptotic pathway, most of the evidence points to an alteration
of mitochondrial function downstream of the Bcl-2 regulatory step. Tumor cells
increase their glycolytic rate by promoting hexokinase localization to the mitochondrial outer membrane and binding to VDAC, where hexokinase uses ATP produced
by oxidative phosphorylation to catalyze the first step of glycolysis [41]. Reconstitution of hexokinase binding to VDAC in liposomes shows that this event also
promotes pore closure, thus this event could regulate apoptosis by modulating the
mitochondrial outer membrane permeability. Consistent with this possibility, forced
disruption of hexokinase association to mitochondria causes cytochrome c release
and apoptosis by a mechanism independent of Bax/Bak and Bcl-2. Conversely,
forced expression of mitochondria-binding hexokinase neutralizes the ability of
activated proapoptotic Bid and Bax to induce apoptosis. AKT prosurvival activities include the upregulation of glucose transporters and maintenance of hexokinase
binding and activity at mitochondria [41]. The resulting upregulation of glucose
metabolism accounts for most of AKT antiapoptotic function, because this activity
can be completely abolished by a hexokinase inhibitor [44].
172
E. Werner
The progressive metabolic dependence of tumors on glycolysis for ATP generation is currently being explored for therapy purposes, revealing glycolysis as an
influential control point for apoptosis. The use of glycolysis inhibitors in tumor cell
lines induces apoptosis even in multidrug-resistant cells and are most effective in
cells with respiratory deficiencies or in those exposed to hypoxia [41, 45].
4.3 Hypoxia
Hypoxia can impose stress on cancer cells due to their reduced access to nutrients and oxygen. However, it can also provide a selective pressure in tumors for the
expansion of variants that have lost their apoptotic potential. Cellular oxygen tension
is sensed by prolyl and asparaginyl hydroxylases, which at normal oxygen levels
target hypoxia inducible factor (HIF) for von Hippel–Lindau dependent degradation. When oxygen tension decreases, HIF accumulates and is able to function as a
transcription factor, inducing the expression of genes involved in both proapoptotic
and antiapoptotic pathways, resulting in the removal of cells that have retained their
apoptotic potential and proliferation of cells that have lost their apoptotic potential [46]. HIF-␣ induces cell death through a p53-dependent mechanism and by
the direct transcriptional induction of proapoptotic BH3-only proteins Noxa, Bnip3,
and Nix [47]. Bnip3 mediates a distinctive form of apoptosis, which is cytochrome
c, AIF, and caspase independent, but involves mitochondrial dysfunction, a mechanism compatible with mitochondrial permeability transition and resembling necrosis in some aspects [48]. Consistent with this finding, high expression levels of the
proapoptotic factors Bnip3 as well as Nix are found in necrotic areas of tumors and
correlate with poor prognosis in cervical cancer, non–small cell lung carcinomas,
and pancreatic cancer.
Tumor hypoxia is an additional selective pressure for p53 inactivation and other
changes leading to apoptosis resistant cells, including Bnip3 silencing. Pancreatic
adenocarcinomas switch off the expression of Bnip3 late in progression by promoter hypermethylation. Additionally, hypoxia is a selective pressure at later stages
of tumorigenesis by promoting apoptosis resistance through antiapoptotic Mcl-1
expression and thereby promoting the selection of highly metastatic Lewis lung carcinoma cells [49].
4.4 Tumor Acidosis
The consequences of increased glucose utilization are lactate production and local
acidification, therefore this will be an attribute of all glycolytic tumors. Relatively
little is known about the molecular effects of environmental acidosis on tumor
physiology. Some evidence suggests that local acidosis could be a barrier to tumor
growth in early carcinogenesis. In vitro studies reveal that acidosis can cause growth
arrest and can alter sensitivity to chemotherapeutic drugs [50]. Environmental acidosis induces p53-dependent cell death in colon adenocarcinoma cells in vitro. Other
How Cancer Cells Escape Death
173
evidence suggests that acidosis exacerbates hypoxia-mediated proapoptotic signals.
This has been observed in normal cardiomyocytes under ischemia, where hypoxia
induces the expression of proapoptotic Bnip3 while acidosis is necessary to accumulate Bnip3 protein in the mitochondria and activate cell death [51].
At later stages of tumor development though, acidosis follows the pattern of all
the other factors reviewed here thus far, correlating with tumor aggressiveness and
the acquisition of proinvasive properties during, for example, the transition from
colon adenomas to invasive cancer and from carcinoma in situ to invasive breast
cancer [52]. Human melanoma cells cultured at low pH exhibit increased invasion
capabilities in vitro and in vivo. Thus, it is foreseeable that acidosis could be a factor
affecting tumor progression. It remains uncertain whether acidosis alters the apoptotic pathway in tumor cells as a factor independent of glycolysis.
4.5 Mitochondria
The pathways leading to cell death or survival converge at the mitochondria for
integration, interpretation, and amplification. Even though mitochondria are at the
center of apoptosis, very little is understood on how mitochondrial function affects
apoptosis and vice versa. Most tumors harbor some level of mutations in mitochondrial DNA, many of them affecting oxidative phosphorylation [53]. Nevertheless,
the role of mitochondrial respiration on apoptosis remains controversial. Some studies indicate that proper Bak-Bax function is dependent on a functional respiratory
chain. Bax function is completely impaired in yeast mutants in any component of the
respiratory chain. These studies contrast with other studies showing that ␳ho0 cells
maintain the capacity to undergo apoptosis in vitro and in vivo, as well as sensitivity
to doxorubicin and growth factor withdrawal [54]. The impact of mitochondrial respiration on apoptosis could be indirect. Acute mitochondrial respiration deficiency
induced by culturing cells in EtBr or treatment with respiratory chain inhibitors is
sufficient to increase the glycolytic rate, to facilitate cell growth in hypoxic conditions, and to reduce susceptibility to several chemotherapeutic agents. This phenotype induced by reducing mitochondrial respiration correlates with increased AKT
activity and can be reversed with inhibitors for this kinase. Thus, this work provides an alternative interpretation to experiments examining respiratory chain function requirement in apoptosis, in addition to suggest that resistance could emerge
from a switch to glycolysis rather than from a decrease in respiratory chain function
per se [55].
Apoptosis can affect mitochondrial function as well. Many experiments hint to
a role for apoptotic players in the control of mitochondrial function. In addition to
cytochrome c participation in the respiratory chain, Bcl-x regulates outer membrane
permeability to anions and inhibits apoptosis by facilitating ADP transport required
to maintain membrane potential. AIF has NADH-dependent oxidase activity, which
is required not only for complex I activity but also for anchorage-independent
growth and tumor formation. Other mitochondrial changes remain poorly studied, although they may be of potential therapeutic interest. Mitochondria from
174
E. Werner
transformed cells and tumors exhibit a higher membrane potential than do normal
cells, correlating with growth rate. Mitochondria isolated from tumors have a higher
relative intrinsic resistance to mitochondrial outer membrane permeability. These
properties have been exploited to concentrate toxic molecules with a delocalized
positive charge in mitochondria and promote apoptosis and mitochondrial dysfunction in transformed cell lines and tumors [56].
5 Concluding Remarks: Implications for Therapy
The wide recognition of the important role played by apoptosis evasion in carcinogenesis has fueled the development of therapies targeted at overcoming this
resistance and thereby increasing cell susceptibility to chemotherapy [57]. After
promising results in preclinical studies, approaches targeting the apoptotic pathway are currently being tested in clinical studies. The results observed illustrate
the advantages and problems associated with these strategies. Initial studies with
G3139/Oblimersen, an antisense oligonucleotide directed at interfering with antiapoptotic Bcl-2 expression, demonstrated a complete lack of effect as a single
therapeutic agent. In combination with fludarabine and cyclophosphamide, which
affect DNA metabolism, however, it caused a limited increase in drug sensitivity in
relapsed chronic lymphocytic leukemia and advanced melanoma patients. Because
drugs directed at proteins involved in apoptosis target a ubiquitous pathway, they
have toxic effects in normal cells, causing neutropenia and thrombocytopenia, as
well as disrupting angiogenesis by altering endothelial cell function [58].
The limited efficacy of Bcl-2 targeting could be explained by several factors.
One possibility is that the results may not correlate with delivery and expression
knock-down efficiency. Another significant factor conferring resistance could be the
presence of other functionally compensating antiapoptotic family members. Thus,
rather than disrupting the function of a single member of the family, interfering with
BH-3 domain activity widens the inhibitory range of the potential drug. The drug
ABT-737 is a potent BH3 peptidomimetic, antagonizing the activity of the antiapoptotic factors Bcl-2, Bcl-xL, and Bcl-w. However, this drug has proved ineffective
when Bcl-2 is phosphorylated or when antiapoptotic Mcl-1 is present [59].
An alternative strategy is to bypass Bcl-2–dependent regulatory mechanism and
target mitochondria directly to trigger apoptosis and release intermembrane effectors. Using this approach, the use of a Smac/Diablo mimetic to activate caspases
directly has been successful in preclinical studies, overcoming resistance induced
by increased Bcl-2 expression, autocrine growth factors, and adhesion in multiple
myeloma [60].
Combination therapies would benefit altogether by considering environmental
and bioenergetic factors affecting the apoptotic pathway, offering additional targets
and aiding in proapoptotic therapy selectivity. In the case of ABT-737, therapeutic
efficacy could be improved by using glucose metabolism inhibitors, which would
cause antiapoptotic Mcl-1 destabilization, as turnover and activity is controlled
by glucose-dependent inhibition of GSK-3 activity [41]. Furthermore, the tumor
How Cancer Cells Escape Death
175
characteristic alterations in metabolism could impart selectivity to apoptosistargeted therapy. Inhibition of glycolysis causes massive apoptosis accentuated in
tumors with mitochondrial respiration deficiency and in hypoxic cells [45]. Reduction of mitochondrial respiration with As2 O3 increases drug-sensitivity as well.
Furthermore, local acidosis could be exploited to activate chemotherapeutic drugs
locally [50]. Thalidomide, a drug known for immunomodulatory effects, disrupts
tumor homeostasis and interaction with the environment. It successfully induced
apoptosis in refractory multiple myeloma cells and decreased the resistance to
chemotherapeutic drugs in vitro, preventing adhesion to bone marrow cells [61].
In prostate cancer, this drug restores interactions with the environment by increasing E-selectin expression [62]. Given these results, thalidomide is currently being
tested in phase II trials for refractory multiple myeloma and solid tumors.
Therapies aimed at modifying the cellular metabolism in order to increase susceptibility to apoptosis address tumor heterogeneity, an additional issue arising in
clinical circumstances and poorly reproduced by mouse models. This is a significant source of variability, accounting in large part for the disappointing inconsistency observed in the results obtained from preclinical and clinical studies, as well
as the difficulties in correlating disease prognosis with apoptotic marker expression
in tumor samples. Tumor models of clonal origin, where oncogenesis is driven by
either the expression of a single oncogene in transgenic mice or by the graft of a
cell line originally isolated as a clone from a tumor, are ideal models to dissect
the role of particular components of a pathway, regulating factors in specific stages
of disease, for evaluation of possible therapeutic targets. However, tumor samples
from patients show high heterogeneity in expression levels of apoptosis markers in
different areas of the same tumor, coexisting with areas of high apoptotic index.
Single cell profiling reveals that heterogeneity determines differential response to
cytokines and drugs and has been proposed to be a key factor for promoting disease
progression in esophageal cancer. Thus, because drugs modulating cell metabolism
do not target specific oncogenic pathways or mutations, they are expected to be more
effective in overcoming tumor heterogeneity as well as in targeting a wider range of
tumors.
Acknowledgments This work was supported by an award from the American Heart Association
and by grant number K22CA127136 from the National Cancer Institute.
References
1. Igney FH, Krammer PH. Death and anti-death: tumour resistance to apoptosis. Nat Rev
Cancer 2002; 2:277–288.
2. Danial NN, Korsmeyer SJ. Cell death: critical control points. Cell 2004; 116:205–219.
3. Adams JM, Cory S. The Bcl-2 apoptotic switch in cancer development and therapy. Oncogene
2007; 26:1324–1337.
4. Zong WX, Lindsten T, Ross AJ, MacGregor GR, Thompson CB. BH3-only proteins that bind
pro-survival Bcl-2 family members fail to induce apoptosis in the absence of Bax and Bak.
Genes Dev 2001; 15:1481–1486.
176
E. Werner
5. Letai A, Bassik MC, Walensky LD, Sorcinelli MD, Weiler S, Korsmeyer SJ. Distinct BH3
domains either sensitize or activate mitochondrial apoptosis, serving as prototype cancer therapeutics. Cancer Cell 2002; 2:183–192.
6. Brenner C, Grimm S. The permeability transition pore complex in cancer cell death. Oncogene
2006; 25:4744–4756.
7. Bouchier-Hayes L, Lartigue L, Newmeyer DD. Mitochondria: pharmacological manipulation
of cell death. J Clin Invest 2005; 115:2640–2647.
8. Chipuk JE, Green DR. Do inducers of apoptosis trigger caspase-independent cell death? Nat
Rev Mol Cell Biol 2005; 6:268–275.
9. Schafer ZT, Kornbluth S. The apoptosome: physiological, developmental, and pathological
modes of regulation. Dev Cell 2006; 10:549–561.
10. Vahsen N, Cande C, Briere JJ, et al. AIF deficiency compromises oxidative phosphorylation.
EMBO J 2004; 23:4679–4689.
11. Boatright KM, Salvesen GS. Mechanisms of caspase activation. Curr Opin Cell Biol 2003;
15:725–731.
12. Riedl SJ, Shi Y. Molecular mechanisms of caspase regulation during apoptosis. Nat Rev Mol
Cell Biol 2004; 5:897–907.
13. Scott CL, Schuler M, Marsden VS, et al. Apaf-1 and caspase-9 do not act as tumor suppressors
in myc-induced lymphomagenesis or mouse embryo fibroblast transformation. J Cell Biol
2004; 164:89–96.
14. Berezovskaya O, Schimmer AD, Glinskii AB, et al. Increased expression of apoptosis
inhibitor protein XIAP contributes to anoikis resistance of circulating human prostate cancer metastasis precursor cells. Cancer Res 2005; 65:2378–2386.
15. Ambrosini G, Adida C, Altieri DC. A novel anti-apoptosis gene, survivin, expressed in cancer
and lymphoma. Nat Med 1997; 3:917–921.
16. Debnath J, Mills KR, Collins NL, Reginato MJ, Muthuswamy SK, Brugge JS. The role of
apoptosis in creating and maintaining luminal space within normal and oncogene-expressing
mammary acini. Cell 2002; 111:29–40.
17. Allen SM, Florell SR, Hanks AN, et al. Survivin expression in mouse skin prevents papilloma
regression and promotes chemical-induced tumor progression. Cancer Res 2003; 63:567–572.
18. Navarro P, Gomez M, Pizarro A, Gamallo C, Quintanilla M, Cano A. A role for the E-cadherin
cell-cell adhesion molecule during tumor progression of mouse epidermal carcinogenesis. J
Cell Biol 1991; 115:517–533.
19. Derksen PW, Liu X, Saridin F, et al. Somatic inactivation of E-cadherin and p53 in mice
leads to metastatic lobular mammary carcinoma through induction of anoikis resistance and
angiogenesis. Cancer Cell 2006; 10:437–449.
20. Boudreau N, Sympson CJ, Werb Z, Bissell MJ. Suppression of ICE and apoptosis in mammary
epithelial cells by extracellular matrix. Science 1995; 267:891–893.
21. Kheradmand F, Werner E, Tremble P, Symons M, Werb Z. Role of Rac1 and oxygen radicals
in collagenase-1 expression induced by cell shape change. Science 1998; 280:898–902.
22. Marconi A, Atzei P, Panza C, et al. FLICE/caspase-8 activation triggers anoikis induced by
beta1-integrin blockade in human keratinocytes. J Cell Sci 2004; 117:5815–5823.
23. Gilmore AP, Metcalfe AD, Romer LH, Streuli CH. Integrin-mediated survival signals regulate
the apoptotic function of Bax through its conformation and subcellular localization. J Cell Biol
2000; 149:431–446.
24. Puthalakath H, Villunger A, O Reilly LA, et al. Bmf: a proapoptotic BH3-only protein regulated by interaction with the myosin V actin motor complex, activated by anoikis. Science
2001; 293:1829–1832.
25. Duxbury MS, Ito H, Zinner MJ, Ashley SW, Whang EE. Focal adhesion kinase gene silencing promotes anoikis and suppresses metastasis of human pancreatic adenocarcinoma cells.
Surgery 2004; 135:555–562.
26. Mazelin L, Bernet A, Bonod-Bidaud C, et al. Netrin-1 controls colorectal tumorigenesis by
regulating apoptosis. Nature 2004; 431:80–184.
How Cancer Cells Escape Death
177
27. Janes SM, Watt FM. Switch from alphavbeta5 to alphavbeta6 integrin expression protects
squamous cell carcinomas from anoikis. J Cell Biol 2004; 166:419–431.
28. Munk Pedersen I, Reed J. Microenvironmental interactions and survival of CLL B-cells. Leuk
Lymphoma 2004; 45:2365–2372.
29. Nefedova Y, Landowski TH, Dalton WS. Bone marrow stromal-derived soluble factors and
direct cell contact contribute to de novo drug resistance of myeloma cells by distinct mechanisms. Leukemia 2003; 17:1175–1182.
30. Sharma SV, Gajowniczek P, Way IP, et al. A common signaling cascade may underlie “addiction” to the Src, BCR-ABL, and EGF receptor oncogenes. Cancer Cell 2006; 10:425–435.
31. Vogelstein B, Kinzler KW. Cancer genes and the pathways they control. Nat Med 2004;
10:789–799.
32. Breitschopf K, Haendeler J, Malchow P, Zeiher AM, Dimmeler S. Posttranslational modification of Bcl-2 facilitates its proteasome-dependent degradation: molecular characterization of
the involved signaling pathway. Mol Cell Biol 2000; 20:1886–1896.
33. McCormick F. Cancer: survival pathways meet their end. Nature 2004; 428:267–269.
34. Zhou XM, Liu Y, Payne G, Lutz RJ, Chittenden T. Growth factors inactivate the cell death
promoter BAD by phosphorylation of its BH3 domain on Ser155. J Biol Chem 2000; 275:
25046–25051.
35. Hammerman PS, Fox CJ, Thompson CB. Beginnings of a signal-transduction pathway for
bioenergetic control of cell survival. Trends Biochem Sci 2004; 29:586–592.
36. Bartkova J, Horejsi Z, Koed K, et al. DNA damage response as a candidate anti-cancer barrier
in early human tumorigenesis. Nature 2005; 434:864–870.
37. Levine AJ. p53, the cellular gatekeeper for growth and division. Cell 1997; 88:323–331.
38. Schmitt CA, Fridman JS, Yang M, Baranov E, Hoffman RM, Lowe SW. Dissecting p53 tumor
suppressor functions in vivo. Cancer Cell 2002; 1:289–298.
39. Xue W, Zender L, Miething C, et al. Senescence and tumour clearance is triggered by p53
restoration in murine liver carcinomas. Nature 2007; 445:656–660.
40. Green DR, Chipuk JE. p53 and metabolism: inside the TIGAR. Cell 2006; 126:30–32.
41. Pelicano H, Martin DS, Xu RH, Huang P. Glycolysis inhibition for anticancer treatment.
Oncogene 2006; 25:4633–4646.
42. Vander Heiden MG, Plas DR, Rathmell JC, Fox CJ, Harris MH, Thompson CB. Growth factors can influence cell growth and survival through effects on glucose metabolism. Mol Cell
Biol 2001; 21:5899–5912.
43. Shim H, Chun YS, Lewis BC, Dang CV. A unique glucose-dependent apoptotic pathway
induced by c-Myc. Proc Natl Acad Sci USA 1998; 95:1511–1516.
44. Gottlob K, Majewski N, Kennedy S, Kandel E, Robey RB, Hay N. Inhibition of early apoptotic events by Akt/PKB is dependent on the first committed step of glycolysis and mitochondrial hexokinase. Genes Dev 2001; 15:1406–1418.
45. Xu RH, Pelicano H, Zhou Y, et al. Inhibition of glycolysis in cancer cells: a novel strategy
to overcome drug resistance associated with mitochondrial respiratory defect and hypoxia.
Cancer Res 2005; 65:613–621.
46. Pouyssegur J, Dayan F, Mazure NM. Hypoxia signalling in cancer and approaches to enforce
tumour regression. Nature 2006; 441:437–443.
47. Kim JY, Ahn HJ, Ryu JH, Suk K, Park JH. BH3-only protein Noxa is a mediator of hypoxic
cell death induced by hypoxia-inducible factor 1alpha. J Exp Med 2004; 199:113–124.
48. Vande Velde C, Cizeau J, Dubik D, et al. BNIP3 and genetic control of necrosis-like
cell death through the mitochondrial permeability transition pore. Mol Cell Biol 2000; 20:
5454–5468.
49. Koshikawa N, Maejima C, Miyazaki K, Nakagawara A, Takenaga K. Hypoxia selects for highmetastatic Lewis lung carcinoma cells overexpressing Mcl-1 and exhibiting reduced apoptotic
potential in solid tumors. Oncogene 2006; 25:917–928.
50. Reichert M, Steinbach JP, Supra P, Weller M. Modulation of growth and radiochemosensitivity
of human malignant glioma cells by acidosis. Cancer 2002; 95:1113–1119.
178
E. Werner
51. Kubasiak LA, Hernandez OM, Bishopric NH, Webster KA. Hypoxia and acidosis activate
cardiac myocyte death through the Bcl-2 family protein BNIP3. Proc Natl Acad Sci USA
2002; 99:12825–12830.
52. Abbey CK, Borowsky AD, McGoldrick ET, et al. In vivo positron-emission tomography imaging of progression and transformation in a mouse model of mammary neoplasia. Proc Natl
Acad Sci USA 2004; 101:11438–11443.
53. Carew JS, Huang P. Mitochondrial defects in cancer. Mol. Cancer 2002; 1:9–21.
54. Jacobson MD, Burne JF, King MP, Miyashita T, Reed JC, Raff MC. Bcl-2 blocks apoptosis in
cells lacking mitochondrial DNA. Nature 1993; 361:365–369.
55. Vander Heiden MG, Chandel NS, Li XX, Schumacker PT, Colombini M, Thompson CB.
Outer mitochondrial membrane permeability can regulate coupled respiration and cell survival. Proc Natl Acad Sci USA 2000; 97:4666–4671.
56. Fantin VR, Berardi MJ, Scorrano L, Korsmeyer SJ, Leder P. A novel mitochondriotoxic small
molecule that selectively inhibits tumor cell growth. Cancer Cell 2002; 2:29–42.
57. Reed JC. Drug insight: cancer therapy strategies based on restoration of endogenous cell death
mechanisms. Nat Clin Pract Oncol 2006; 3:388–398.
58. Zeitlin BD, Joo E, Dong Z, et al. Antiangiogenic effect of TW37, a small-molecule inhibitor
of Bcl-2. Cancer Res 2006; 66:8698–8706.
59. Konopleva M, Contractor R, Tsao T, et al. Mechanisms of apoptosis sensitivity and resistance
to the BH3 mimetic ABT-737 in acute myeloid leukemia. Cancer Cell 2006; 10:375–388.
60. Chauhan D, Neri P, Velankar M, et al. Targeting mitochondrial factor Smac/DIABLO as therapy for multiple myeloma (MM). Blood 2007; 109:1220–1227.
61. Corso A, Ferretti E, Lunghi M, et al. Zoledronic acid down-regulates adhesion molecules
of bone marrow stromal cells in multiple myeloma: a possible mechanism for its antitumor
effect. Cancer 2005; 104:118–125.
62. Efstathiou E, Troncoso P, Wen S, et al. Initial modulation of the tumor microenvironment
accounts for thalidomide activity in prostate cancer. Clin Cancer Res 2007; 13:1224–1231.
Index
A
Abate, C., 64
Abbey, C. K., 173
Abu-Hamad, S., 12
Acetoin, 106, 107
Acin-Perez, R., 128
Acker, T., 135
Aconitase in Krebs cycle, 4
Adams, J. M., 36, 164, 166
Adebanjo, O. A., 63
Adida, C., 167
Agani, F., 82
Ahern, K. G., 106, 113
Ahn, H. J., 172
Ainscow, E. K., 115
Akimoto, M., 59, 123
Akt/PBK (protein kinase B)
pathway, 8
Alam, N. A., 25, 59
Alarcon, R., 138
Alazard, N., 58, 62
Alderson, M. R., 149
Aldolase (ALD), 84, 85
Alessi, D. R., 12
Allan, S., 94
Allen, S. M., 168
Almodovar, C. R., 114
Alonso, A. M., 6, 10
Alonso, C., 30
Alquier, T., 12
Altenberg, B., 6
Altieri, D. C., 167
Ambrosini, G., 167
Amir, S., 136
Amstad, P., 64
Amuthan, G., 48–50, 58, 61–63, 67, 119
Ananadatheerthavarada, H. K., 48–50, 58,
61–63, 67
Anbazhagan, R., 62
Anderson, S., 121
Andersson, U., 61
Angiogenesis, 37–39
Antiapoptotic genes, 66–67
Antioxidant
antioxidant response element
(ARE), 38
defense pathway, 65
enzyme levels, 105
Aoyama, N., 7
Apoptosis, 36, 47, 162–163
Bcl-2 family proteins, 164–165
caspases and, 166
Appel, L. J., 2
Apse, K., 97
Armour, S. M., 24
Arnold, R. S., 35, 38
Arora-Kuruganti, P., 63
Aryl hydrocarbon nuclear translocator
(ARNT), 135
Asham, A. M., 134
Ashley, S. W., 169
Aslan, M., 27
Astuti, D., 23, 59, 136
Attardi, G., 60
Atzei, P., 168
Augenlicht, L. H., 126
Avadhani, N. G., 48–50, 58, 61–63, 67, 119
Azoulay-Zohar, H., 12
B
Babior, B. M., 20
Bach, D., 66
Bader, S., 139
Bae, S. H., 24, 80, 81
Bae, Y. S., 34
Baeuerle, P. A., 64
Baggett, B. K., 8
Baggetto, L. G., 9, 106, 107
179
180
Baldi, P., 25, 27, 47, 127
Bale, A. E., 46
Ballard, J. W. O., 108
Bamezai, R. N., 128
Bandelt, H. J., 22, 47, 126–128
Bankier, A. T., 121
Baranov, E., 171
Barrell, B. G., 121
Barrett, J. N., 60, 61
Bartkova, J., 170
Bartnik, E., 121, 123, 126
Bassik, M. C., 165
Battersby, B. J., 128
Bauer, C., 78
Bauer, D. E., 8, 79, 113
Baumann, A. K., 9, 58, 64, 127
Bayascas, J. R., 12
Baysal, B. E., 9, 20, 23, 59
Beach, D., 94, 95
Beasley, N. J., 137
Bellamy, W. T., 29
Bell, E. L., 38
Benit, P., 24, 25
Benizri, E., 77
Bensaad, K., 9
Ben-Shlomo, I., 115, 116
Berardi, M. J., 174
Berchner-Pfannschmidt, U., 75, 77
Berezovskaya, O., 167
Berman, S. B., 47
Bernard, D., 132
Bernet, A., 169
Berra, E., 77
Bhat, A. K., 128
Bhat-Nakshatri, P., 64
Bhujwalla, Z. M., 137
Bhushan, A., 148
Bianchi, M. S., 124, 125
Bianchini, F., 107
Bianchi, N. O., 124, 125
Bindra, R. S., 134
Birch-Machin, M., 123
Birnbaum, M. J., 7
Birrer, M. J., 64
Bishopric, N. H., 114, 173
Bissell, M. J., 168
Biswas, G., 48–50, 58, 61–63, 67, 119
Bi, X., 11
Blachly-Dyson, E., 11
Blanco-Rivero, A., 2, 9–11
Blasco, M. A., 94
Bluyssen, H. A., 140
Boatright, K. M., 166
Index
Bodyak, N. D., 57, 58, 125, 126
Bond, J. D., 63
Bonhoeffer, S., 2, 5, 9
Bonicalzi, M. E., 139
Bonni, A., 41
Bonod-Bidaud, C., 169
Bonora, E., 127
Booker, L. M., 128
Borisenko,G. G., 107
Borowsky, A. D., 173
Bottoni, P., 50, 51
Bouchier-Hayes, L., 165
Boudreau, N., 168
Bouillaud, F., 21
Boulianne, G. L., 96
Bourgeron, T., 22, 25
Bove, K., 38
Bowden, G. T., 57, 64, 65
Boyer, P. D., 3
Bracale, R., 107
Braithwaite, K. L., 92
Brand, M. D., 115
Brandon, M., 25, 27, 47, 127
Brands, M., 2
Braunstein, L. D., 97
Breitschopf, K., 170
Brenner, C., 165
Bretscher, P. A., 149
Briehl, M. M., 29
Briere, J. J., 12, 20, 23–25, 166
Brigelius-Flohe, R., 64
Briggs, J., 140
Briones, P., 66
Bromley, L., 63
Brown, D. T., 124
Brown, S. J., 46
Brundtland, G. H., 148
Brune, B., 138
Br¨une, B., 79
Brunelle, J. K., 38
Brunet, A., 41
Bryk, J., 128
Buecher, T., 22
Bulavin, D. V., 11
Bunn, H. F., 136
Burdon, R. H., 63
Buricchi, F., 37
Burk, D., 5
Burne, J. F., 173
Bustamante, E., 8, 27
Butow, R. A., 67, 68
Butts, B., 65
Buzzai, M., 8, 79
Index
C
Caballero, O. L., 57
Caenorhabditis elegans and mev1 mutant, 23
Cairns, R. A., 10
Calcium signaling pathway, 62, 63
Calderwood, S. K., 48
Calle, E. E., 107
Camenisch, G., 138
Cameron, G., 133
Camley, R. E., 152
Campanella, C., 48
Campisi, J., 95, 108
Cancer, 46
apoptotic pathway in, 166–167
cells, 145
Achilles’ heel of, 29
AKT activation in, 78–79
bioenergetic signature of, 9
HIF-1α expression in, 88
Notch-3 and carbonic anhydrase IX
(CA-IX), 42
PTENoxidative inactivation and AKT
signaling pathway, 41
retrograde signaling, 60–67
Warburg effect in, 8, 46
Warburg phenotype of, 10–11
COXI mtDNA mutations, 58
drug treatment and Fas-induced tumor cell
death, 150–157
fumarate hydratase (FH) mutations, 88
G3139/Oblimersen studies with, 174
HKII transcription, 83
homoplasmy and heteroplasmy, 125
immune
response in, 147–148
system and, 149
metabolic changes in cell, survival effects
glycolysis, 171–172
hypoxia, 172
mitochondria and, 173–174
p53 tumor suppressor, 170–171
tumor acidosis, 172–173
mitochondrial genome instability (mtGI)
in, 124–125
OXPHOS
defects in, 56
dysfunction and, 66
Smac/Diablo mimetic use in, 174–175
surveillance mechanism
cell-cell interactions and tissue
architecture, 167–169
oncogene activation and anti-oncogene
silencing, 169–170
181
thalidomide drug for, 175
therapeutic possibilities, 158–159
Cande, C., 12, 20, 166
Cano, A., 168
Cappello, F., 48
Carcinogenesis
aerobic respiration and energy generation,
104
calorie restriction inhibition
energy expenditure and intake, 108
metabolic disorders, 107
UDPGlcNAc decreased levels, 109
chronic infection and
signaling molecules, 110
diversion of pyruvate mitochondria
respiration (PDH) and fermentation
(PDC), 106
TCA cycle, 106
electron transport modulation
NADH/cyt-C electron transport
pathway, 110
glycolysis involvement in, 93
immortalization and transformation, 91–92
induced apoptosis, 111–112
metabolic modulation of, 103
organelle and cell membrane
perturbations, 107
OXPHOS phenotype and apoptosis
ATP synthase and suppression of, 104
reactive oxygen species and
cell-signaling pathways, 105
mitogenic signals, 105
Warburg model, 113–115
Cardile, A., 107
Cardiolipin (CL), 107
Carew, J. S., 9, 120, 122–125, 173
Carling, D., 12
Carmeliet, P., 138
Carney, D., 126
Carothers, A. D., 126, 128
Carotid body PGL, 23
Carracedo, A., 22, 47, 126–128
Carter, M. J., 107
Casado, E., 9–11
Casimiro, M. C., 11
Caspases and cellular proteins, 36
Castilla, J., 30
C2C12 rhabdomyosarcoma, ryanodine
receptor-1 (RyR-1) and-2 (RyR-2)
genes, 63
Cellular respiration
tumor suppressor genes
p53, 137–138
182
Cellular respiration (cont.)
von Hippel–Lindau syndrome, 138–140
and Warburg hypothesis, 132–133
Cepeda, V., 30
Cerutti, P., 64
Cerutti, P. A., 63
Chan, D. C., 66
Chandel, N. S., 173
Chandra, D., 48
Chang, G. W., 136
Chang, H. J., 10
Chang, I., 11
Chang, T. S., 34
Chan, T. L., 57
Chauhan, D., 174
Chemodectoma, see Carotid body PGL
Chemotherapeutic-resistant tumor cells
sequester cytosolic calcium
(Cacyt ), 48
Chen, E. I., 52
Chen, G., 6, 10
Cheng, G., 38
Cheng, M. L., 97
Cheng, W. C., 47
Chen, H., 66
Chen, X., 22
Chen, Z., 11
Chen, Z. J., 65, 110
Cheung, A. N., 57
Cheung, H. S., 61
Chiarle, R., 23
Chiarugi, P., 37
Chieco, P., 140
Chilov, D., 138
Chinnery, P. F., 124–126, 128
Chipuk, J. E., 171
Chi, S. M., 135
Chittenden, T., 170
Cho, E. J., 140
Choi, J., 100
Choi, K. Y., 107
Chomyn, A., 66
Chretien, D., 22, 25
Chu, G., 42
Chun, Y. S., 171
Chu, S. C., 41
Cinalli, R. M., 8
Ciocca, D. R., 48
Cisplatin-resistant ovarian carcinoma, 38
Cizeau, J., 172
Clayton, D. A., 22
Clear cell renal carcinoma (CCRC) cells, 81
Clifford, S. C., 136, 139
Index
Cockman, M. E., 139
Cohen, A., 110
Cohen, C., 7
Cohn, M., 149
Colburn, N. H., 65
Coller, H. A., 57, 58, 125, 126
Colombini, M., 173
Combs, C. A., 42
Connor, K. M., 37, 38, 40
Contractor, R., 174
Copeland, W. C., 121, 122, 124
Corral-Debrinski, M., 22
Corso, A., 175
Cory, S., 164, 166
Coughlin, S. S., 107
Couplan, E., 21
Cozzi, V., 107
Crawford, D., 64
Crisp, W., 152
Crosby, M. E., 134
Cucuianu, A., 109
Cuezva, J. M., 2, 6, 7, 9–12
Curran, T., 64
Cycloheximide (CHX), 36
Cytochrome c oxidase (COX), 9
Czarnecka, A. M., 48, 121, 123, 126
D
Dachs, G. U., 8
Dalgard, C. L., 8
Dallol, A., 23, 59
Dalton, W. S., 169
Dang, C. V., 7, 10, 46, 92, 134, 171
Dang, G., 7
Danial, N. N., 12, 164, 166
Darin, N., 22
Darvishi, K., 128
Dasgupta, J., 37
Dayan, F., 172
Debatin, K. M., 30
Decorps, M., 7
de Heredia, M. L., 6, 7, 9–11
Dellian, M., 134
Delsite, R., 62
De, Marzo, A. M., 80
Deml, E., 9
Demont, J., 9, 10, 58, 59
Denko, N. C., 10, 134
de Paulsen, N., 139
De Pinho, R. A., 94
Derksen, P. W., 168
Desbaillets, I., 138
Desbarats, J., 149
Index
Desvergne, B., 51
Dey, R., 11, 67
Diala, E. S., 66
Dickinson, D., 96
Dictyostelium discoideum and phosphoinositide (3, 4, 5) 3-phosphate
(PtdIns[3, 4, 5]P3), 40
Dietrich, A., 22
Dimmeler, S., 170
Ditsworth, D., 113
Dolde, C., 7, 93
Domann, F. E., 64
Domann, F. E. Jr., 64
Don, A. S., 123
Dong, Z., 64, 174
Dor, Y., 138
Douglas-Jones, A., 61
Drago, I., 62
Droge, W., 47
Drosophila simulans, 108
Dubik, D., 172
Dujon, B., 22
Dukes, I. D., 115
Du, M., 10, 113
Dunn, G. P., 110
Dunning, S. P., 136
Du, X., 113
Duxbury, M. S., 169
E
Earl, A., 66
Ebert, B. L., 8
Edelstein, D., 113
Edens, W. A., 38
Efstathiou, E., 175
Egea, G., 11
Electron transport chain (ETC)
cell death and proliferation, 28
in mammalian cell, 56
modulation, 110–111
El Ghouzzi, V., 25
Elia, A. J., 96
Elson, J., 125
Elstrom, R. L., 8, 79
Endo, T., 61
Endothelial PAS-domain protein
1 (EPAS1), 79
Energy-deprived HepG2 hepatoma cells,
mitochondrial biogenesis and
OXPHOS, 108
Enolase (ENO), 85
Enriquez, J. A., 119, 121, 125
Epstein, C. B., 68, 137
183
Erlacher, M., 137
Ernest, I., 135
Erythropoietin (EPO) and VEGF, role in
hypoxia, 80
E3 ubiquitin ligases and HIF-1α, 78
Eukaryotic peroxiredoxins (PRXs), 35
Evan, G. I., 150
Evans, I. H., 66
F
Fabregat, I., 10–12
Faddy, H. M., 62
Faller, D. V., 135
Fang, H. M., 8, 133, 135, 136
Fang, W. G., 135
Fan, M., 65
Fantin, V. R., 46, 49, 51, 174
Favier, J., 23–25
Felty, Q., 50
Fernandez, A., 135
Fernandez, P. L., 6, 9, 11
Fernandez-Silva, P., 119, 121, 125, 128
Ferrans, V. J., 34
Ferrell, R. E., 9, 20, 23, 59
Ferretti, E., 175
Fibrate/thiazolidinedione-induced
differentiation, 51
Finch, J. S., 64
Finkel, T., 34, 42
Fink, K., 124
Firth, J. D., 8
Flier, J. S., 7
Fliss, M. S., 57
Flohe, L., 64
Floodgate hypothesis, 35
Florell, S. R., 168
Floyd, R. A., 35
Foekens, J. A., 140
Foker, J., 11
Folkman, J., 134
Fontana, L., 10
Fontana, L. A., 134
Foo, T. W., 11
Forbes, R. A., 82, 136
Forkhead transcription factors (FoxO), 41–42
Fornace, A. J. Jr., 11
Fox, C. J., 8, 170, 171
Franc, B., 59
Francis, M. J., 63
Frataxin and Fe-S cluster biogenesis, 11
Freedman, B. D., 63
Freedman, D. A., 137
French, S., 42
184
Fridman, J. S., 171
Fuertes, M. A., 30
Fuhrer, J. P., 66
Fulda, S., 30
G
Gabrielson, E., 62
Gajewski, C. D., 64
Gajowniczek, P., 169
Galluzzi, L., 30
Gamallo, C., 168
Gambhir, S. S., 7
GAPDH overexpression and human prostate
adenocarcinoma cells (cell line
LNCap), 85
Garber, K., 46, 137
Gar´ca-Gar´ca, E., 6, 7, 9, 10
Garrido, C., 20
Gasparre, G., 127
Gassmann, M., 78, 138
Gatenby, R. A., 7, 46, 50, 93, 115
Gatter, K. C., 8
Gauthier-Rouviere, C., 135
Gebbink, M. F., 140
Genes encoding complex II mutations, tumor
and cancer formation, 23
Geromel, V., 22, 25
Giacomello, M., 62
Giardia lamblia, 4
Giardina, B., 50
Giatromanolaki, A., 8
Gil, J., 96, 97, 132
Gill, V., 63
Gillies, R. J., 7, 8, 46, 50, 93, 115
Gilmore, A. P., 169
Gimenez-Roqueplo, A. P., 23, 24
Ginkgo biloba, proapoptotic molecules
expression, 37
Ginouv´es, A., 77
Giorgio, M., 42
Giovannini, C., 42
Glazer, P. M., 134
Gleadle, J. M., 8
Glinskii, A. B., 167
Glucose-6-phosphate dehydrogenase (G6PD)
activity on cell proliferation, 97
Glucose phosphate isomerase (GPI), 83
Glucose transporters (GLUT1 to GLUT5), 87
Glyceraldehyde phosphate dehydrogenase
(GAPDH), 85
Glycolysis, 2
aldolase (ALD), 84–85
apoptotic pathways and, 114
Index
ATP levels and OXPHOS, 111
cell cycle and apoptotic pathway,
regulation, 112
and cellular senescence, 94–95
enolase (ENO), 85
glucose phosphate isomerase (GPI), 83
glyceraldehyde phosphate dehydrogenase
(GAPDH), 85
glycolysis-dependent phenotype
of malignancies, 8
hexokinases (HKs), 82–83
hypoxia inducible factor 1 alpha
(HIF-1α), 8
induction of hypoxia path by succinate
through stabilization of, 26
hypoxic conditions and, 92–93
lactate dehydrogenase and lactate carrier
MCT4, 86–87
LMW-PTP activity and, 37
malignant growth and, 5
Nox1-dependent generation of H2 O2 , 38
2-OG dioxygenase, 75–76
oncogene and, 93–94
organisms life span and
calorie restriction, 98–99
Sir2 as longevity gene in Yeast, 97–98
phosphofructokinase 1 (PFK-1), 84
phosphoglycerate kinase (PGK), 85
pyruvate dehydrogenase (PDH), 86
pyruvate kinase (PK), 86
as radical scavenger, 97
regulatory mechanism of
p53 and, 99–100
and PGM, 100–101
Sp1 and Bnip3, 106
TIGAR acts as inhibitor of, 9
Glycolytic malignant cells, 106
Gnarra, J., 139
Gnarra, J. R., 139
Goeddel, D. V., 36
Goldberg, M. A., 136
Goldblatt, H., 133
Goldin, A., 150
Golik, P., 121, 123, 126, 128
Golstein, P., 20, 149
Gomez, M., 168
Gordan, J. D., 24, 113
Gorgoglione, V., 110
Goswami, P. C., 36
Gottlieb, E., 47, 121
Gottlieb, R. A., 20
Gottlob, K., 171
Goulet, R. J. Jr., 64
Index
Govindarajan, B., 7, 8
Graeber, T. G., 137
Gramlich, J. L., 41
Gramm, C. F., 12
Greco, M., 60
Greenawalt, J. W., 57
Greenblatt, M. S., 137
Green, D. R., 4, 19, 20, 150, 171
Greim, H., 9
Greulich, K. O., 6
Grimm, S., 165
Gross, A., 11
Groszer, M., 10
Groulx, I., 139
Grover-McKay, M., 7
Guan, K. L., 12
Guarente, L., 98
Guppy, M., 5
Gupta, A., 57, 64, 65
Gutman, M., 22
Gu, W., 132
Gu, Y., 110
Guyetant, S., 59
Gyllensten, U., 127
H
Habano, W., 9
Habermacher, G. M., 128
Haendeler, J., 170
Hagler, J., 65
Hale, W. T., 68
Hammerman, P. S., 8, 170
Hammond, E., 138
Hanahan, D., 11, 92, 134
Hanks, A. N., 168
Hannan, R. L., 135
Hannon, G. J., 94
Haraguchi, K., 61
Hardie, D. G., 12
Harman, D., 94
Harper, M-E., 149, 150, 154–158
Harris, A. L., 8, 134, 140
Harris, M. H., 11, 108, 171
Hashida, M., 27
Haspel, H. C., 7
Hayflick, L., 94
Hay, N., 171
Heat shock proteins (HSPs), 48
Heerdt, B. G., 126
Heese, C., 9
Helicobacter pylori, 109
Helmlinger, G., 134
Hendrix, M. J., 7
185
Hepatitis B and Epstein-Barr viruses, 109
Herbert, J. M., 138
Herman, J. G., 139
Hernandez, O. M., 114, 173
Herrero-Jimenez, P., 57, 58, 125, 126
Herrmann, E. C., 9
Herrmann, P. C., 9
Herrnstadt, C., 47
Hershman, J. M., 61
Hervouet, E., 9, 10
Hewel, J., 52
Hexokinase II, 48
Hexokinases (HKs), 82–83
Hexosamine biosynthetic pathway (HBP)
nutrient uptake pathways, 105, 108
O-GlcNAc glycosylation, 105–106
Hickey, M. M., 24
Hickman, J. A., 51
Higuchi, T., 9
Hilliker, A. J., 96
Hirabayashi, Y., 29
Hirota, K., 132
Hoberman, H. D., 9
Hock, J. B., 104
Hoffman, M. A., 139
Hoffman, R. M., 171
Hogg, P. J., 123
Ho, H. Y., 97
Hollstein, M., 137
Hooper, L., 136
H¨opfl, G., 78
Horejsi, Z., 170
Horiuchi, S., 9
Horton, R., 126, 128
Horvath, J. C., 150
Howell, N., 47
Hruban, Z., 66
HSP70 overexpression, 48
Huang, C., 64
Huang, L. E., 134
Huang, P., 9, 120, 122–126, 171, 173
Hudson, K. M., 134
Human osteosarcoma cells, 104
Human placenta choriocarcinoma cell line
BeWo (JCRB911) and GLUT1
mRNA, 87
Human spontaneous cervical cancer
cells (SiHA) and GAPDH
overexpression, 85
Huntly, B. J., 42
Hwang, P. M., 10
Hypoxia, 38, 107, 109, 134–135
hypoxia response elements (HREs), 80
186
Hypoxia (cont.)
hypoxic cells, studies of GLUT mRNA
expression in, 87
Hypoxia inducible factor 1 α (HIF-1 α)
in cancer, 80–81
and glycolytic pathway, 81–82
ectopic expression of, 94
glucose transporters and, 87–88
HRE and, 135–136
nonhypoxic stimuli, induction by, 80
regulation, 75
acetyl transferase, 79
MAPK and PI3K pathways, 78
polyubiquitination, 77–78
stabilization in normoxia condition,
76–77
structure of, 75
transactivation domains and, 74
TCA intermediates upregulation and
carcinogenesis, 88
as therapeutic target, 88
I
Ichikawa, M., 59, 123
Ido, Y., 113
Igney, F. H., 162
Immortalization, 92
enhanced glycolysis
senescence and, 94–95
See also Carcinogenesis
Inghirami, G., 23
Ingman, M., 127
Ingram, V. M., 63
Inoki, K., 12
Irani, K., 57
Iron sulfur (Fe-S) clusters, mitochondrial
enzymes, 4
Isidoro, A., 6, 9–11
Isken, F., 11
Israelson, A., 12
Itahana, K., 96
Itahana, Y., 96
Ito, H., 169
Ivan, M., 139
Ivanovska, I., 47
Iyer, N. V., 82
Izquierdo, J. M., 6, 11
J
Jaattela, M., 4
Jacks, T., 137
Jacobs, H. T., 22
Jacobson, M. D., 173
Jacques, C., 60, 61
Index
Jagiello, G., 22
Jain, R. K., 134
Janes, S. M., 169
Jazwinski, S. M., 68
Jeong, J. W., 24, 80, 81
Jessie, B. C., 128
Jewell, U. R., 78
Jiang, B. H., 135
Jiang, J., 107
Jiang, W. G., 61
Jiao, J., 10
Johnson, F. M., 121, 122, 124
Johnson, J. H., 115
Jones, R. G., 12
Joo, E., 174
June, C. H., 149
Ju, X., 11
K
Kaaks, R., 107
Kachhap, S., 10, 62, 127, 128
Kadhom, N., 25
Kaelin, W. G., 24, 139
Kaelin, W. G. Jr., 23, 139
Kagan, V. E., 107
Kahn, B. B., 12
Kanamori, T., 58, 123, 127
Kandel, E., 171
Kang, J. G., 10, 100, 132
Kang, S. W., 34
Kappler, J., 149
Karin, M., 64, 65
Karplus, P. A., 35
Kaschten, B. J., 7
Kataoka, T., 149
Kearns-Sayre syndrome (KSS), 124
Kennedy, S., 171
Khaleque, M. A., 48
Kheradmand, F., 168
Khrapko, K., 57, 58, 125, 126
Kilo, C., 113
Kim, H., 140
Kim, J. H., 41, 79
Kim, J. W., 10, 46, 94
Kim, J. Y., 11, 172
Kim, K. W., 24, 80, 81
Kim, M. K., 107
Kim, S., 7
Kim, S. H., 24, 79–81
Kim, S. T., 140
Kim, W. Y., 139
Kim, Y. H., 11
King, A., 47
Index
King, M. P., 173
Kinzler, K. W., 137, 169
Kirkinezos, I. G., 124
Kispal, G., 11
Klatt, P., 100
Klausner, R. D., 4
Klco, J., 139
Klco, J. M., 139
Klein-Szanto, A., 58, 61, 62, 119
Klingenberg, M., 22
Kniffin, C. L., 8
Knudson, A. G. Jr., 139
Koch, C. J., 134
Koed, K., 170
Kofler, B., 124
Kohl, R., 138
Kollipara, R., 42
Kol, S., 115, 116
Kondoh, H., 96, 97, 100, 132
Kondo, K., 139
Konopleva, M., 174
Koocheckpour, S., 140
Korsmeyer, S. J., 164–166, 174
Koshikawa, N., 172
Kotch, L. E., 82
Koukourakis, M. I., 8
Koumenis, C., 138
Kourembanas, S., 135
Kovacs, G., 57
Ko, Y. H., 27, 48
Krajewska, M., 6, 7, 9–11
Krammer, P. H., 162
Krebs cycle, 3, 107
Krebs, H., 6, 133
Krippner, A., 20
Kroemer, G., 19, 20, 25, 30
Kron, S. J., 11, 108
Krueger, J. S., 52
Kruse, J. P., 132
Kubasiak, L. A., 114, 173
Kubek, S., 12
Kuchroo, V. K., 149
Kuzmin, I., 140
Kvietikova, I., 78
Kwei, K. A., 65
Kwong, J. Q., 67
L
Lactate dehydrogenase A (LDH-A)
expression, 52
and lactate carrier MCT4, 86, 87
Lambeth, J. D., 38
Lamb, N., 135
187
Landowski, T. H., 150, 169
Lanier, L. L., 149
Larochette, N., 30
Larsson, N. G., 22
Larsson, R., 64
Lartigue, L., 165
Latif, F., 23, 59, 139
Laughner, E., 80
Leber hereditary optic neuropathy
(LHON), 124
Lechago, J., 7
Lechago, L. V., 7
Lechpammer, M., 139
Ledbetter, J. A., 149
Leder, P., 46, 49, 51, 174
Lee, F. S., 65
Lee, H., 106
Lee, J. W., 24, 80, 81
Lee, S., 139
Lee, S. M., 140
Lee, S. R., 34
Lee, S. Y., 107
Lee, T. Y., 11
Lee, W. C., 107
Lehninger, A. L., 10, 106
Leon-Avila, G., 4
Letai, A., 165
Leukemia, 109
Leung, S. W., 135
Levine, A. J., 137, 170
Levy, J. P., 64
Lewis, B. C., 7, 93, 171
Lichtenstein, A. H., 2
Lien, A. D., 8
Lightowlers, R. N., 22
Li, G. X., 29
Li, J. J., 65
Lill, R., 4, 11
Lim, A. L., 10
Lin, A. W., 95
Lin, Q., 11
Lin, S. J., 98
Lincoln, D. W., 38
Lindsten, T., 165
Linsley, P. S., 149
Lipid signaling molecules and migratory
phenotype regulation, 40
Liu, J., 23
Liu, J. W., 48
Liu, V. W., 57
Liu, X., 168
Liu, Y., 170
Liu, Z., 64
188
Li, X. D., 110
Li, X. X., 173
Li, Y., 57, 125, 126
Li, Z., 11
Ljungberg, B., 57
Lleonart, M. E., 96, 97, 100, 132
Lodish, H. F., 7
Lofromento, N. E., 110
Lolkema, M. P., 140
Lonergan, K. M., 140
L´opez de Heredia, M., 6, 11
L´opez-R´os, F., 6, 7, 9, 10
Lorenc, A., 128
Lorenz, M., 96
Los, M., 140
Lott, M. T., 120, 127
Lou, Y. R., 29
Lowe, S. W., 95, 171
Lucchesi, P. A., 63
Lu, F. J., 97
Lu, H., 8, 82, 136
Luis, A. M., 11
Lunghi, M., 175
Luo, Y., 63
Lutz, R. J., 170
Lu, X., 138
Lu, Y., 11
Lu, Y. P., 29
Lynch, M., 121
M
Mabjeesh, N. J., 136
Macaulay, V., 22, 47, 126–128
MacGregor, G. R., 165
MacKenzie, E. D., 24, 25
Maejima, C., 172
Maher, E. R., 136
Mahon, P. C., 132
Majewski, N., 171
Malchow, P., 170
Malthiery, Y., 59
Mammalian mitochondrial oxidative
phosphorylation (OXPHOS), 4, 55
Bcl-2 expression and, 67
defects in cancer
cellular adaptations and, 59–60
functional significance of, 57–59
dysfunction
mtDNA mutations and, 58
tumor-specific marker genes, 61–62
gene expression and transcription, 106
mitochondrial-nuclear intergenomic
signaling in, 59
Index
phenotype and apoptosis, 104
representation of, 56
Manfredi, G., 67
Manganese-dependent superoxide dismutase
(MnSOD) overexpression and cell
proliferation, 28
Maniatis, T., 65
Manka, D., 106
Mansel, R. E., 61
Mansouri, J., 41
MAPK pathways, 64
Maraver, A., 100
Marconi, A., 168
Marin-Hernandez, A., 46
Marquardt, C., 11
Marrack, P., 149
Marsden, P. A., 135
Marsden, V. S., 167
Martin, D. A., 64
Martindale, D., 149
Martin, D. S., 171
Martinez-Diez, M., 10–12
Martinez, M., 6, 9, 11
Martorana, G. E., 50, 51
Marzulli, D., 110
Matheu, A., 100
Mathupala, S. P., 9, 27, 48
Matoba, S., 10, 100, 132
Matsumura, T., 113
Matsuno-Yagi, A., 20
Matsuyama, S., 11
Matthew, C. K., 106, 113
Mattiazzi, M., 64
Mavligit, G. M., 149
Ma, W., 10
Maxwell, P. H., 8, 136, 140
Mayr, J. A., 124
Mazelin, L., 169
Mazure, N. M., 172
McAndrew, D., 62
McCormick, F., 170
McCurrach, M. E., 95
McFate, T., 8
McGoldrick, E. T., 173
McKnight, S. L., 11
McKusick, V. A., 8
McQuillan, L. P., 135
Mechta-Grigoriou, F., 80
Meierhofer, D., 124
Melamede, R. J., 152
Melendez, J. A., 38, 40
Melvin, R. G., 108
Menon, S. G., 36
Index
Mertz, R. J., 115
Merzak, A., 140
Metcalfe, A. D., 169
Metzen, E., 75, 77
Miceli, M. V., 68
Michael, E. M., 124
Michalak, E., 137
Michalik, L., 51
Michel, G., 135
Miething, C., 171
Migliaccio, E., 42
Millhorn, D. E., 106
Minami, R., 7
Minet, E., 135
Mirebeau, D., 60, 61
Mitochondria
activities in, 4
and apoptosis, 29
cancer, 20
cell immune system, 110
cellular respiration and dedifferentiation,
49–52
DNA deletions and mutations, 110
genome, 121–124
H2 O2 -dependent transcription of MMP-1
transcription, 40–41
H2 O2 , generation of, 37
metabolism in, 147
mitochondrial-derived ROS, 34
angiogenesis, 37–39
cancer stem cells, 41–42
and cell cycle, 35–36
cell survival, 36–37
invasion/migration, 39–41
mitochondrial DNA (mtDNA)
mutations, 47, 57–58
mitochondrial encephalomyopathy lactic
acidosis, and strokelike episodes
(MELAS), 124
mitochondrial HSP-mediated, 51
mitochondrial transcription factor A
(TFAM), 61
optic atrophy 1 (Opa 1), 66
outer membrane permeability, 165–166
OXPHOS and respiration, 5–6, 10
peroxisome proliferator activated receptor
(PPAR)-α–mediated effects in
rodents, 51
p66shc regulation and, 42
respiration and ROS production, 106
specific phospholipid cardiolipin (CL)
complexation, 107
superoxides role, 27
189
type II hexokinase, interaction between, 24
voltage-dependent anion channel
(VDAC), 27
Mitofusin 2 (Mfn 2) downregulation, 66
Miyashita, T., 173
Miyazaki, K., 172
MnSOD-overexpressing MCF-7 cells, 65
Modica-Napolitano, J. S., 47
Mohyeldin, A., 8
Molecular homeostasis and genetic instability,
133–134
Montanaro, L., 140
Monteith, G. R., 62
Montoya, J., 119, 121, 125
Mookerjee, B., 137
Moorhead, P. S., 94
Moraes, C. T., 11, 60, 61, 67, 124
Moreno-Loshuertos, R., 128
Moreno-Sanchez, R., 46
Morris, H. P., 8, 57
Morrissey, C., 140
Mueckler, M. M., 7
Muehlematter, D., 64
M¨uhlenhoff, U., 4, 11
Muller, U., 23, 59
Munk Pedersen, I., 169
Munnich, A., 22, 27
Murad, E., 35
Murrell, G. A., 63
Muse, K. E., 65
Myoclonic epilepsy and ragged-red fibers
(MERRF), 124
N
NADPH oxidase family members (Nox1-5), 35
Nagata, S., 149
Nagy, A., 57
Nakagawara, A., 172
Nakamura, E., 23, 139
Nakamura, S., 9
Nakashima, Y., 96, 100
Nakshatri, H., 64
Nass, M. M., 66
Navarro, P., 168
Nefedova, Y., 169
Negelein, E., 5, 20
Neilson, A., 9
Nejfelt, M. K., 135
Nekhaeva, E., 57, 58, 125, 126
Nelson, K. K., 38, 40, 41
Nemoto, S., 34, 42
Neri, P., 174
Neumann, H. P., 9
190
Neurogenic muscle weakness, ataxia and
retinitis pigmentosa (NARP), 124
Newell, M. K., 149, 152
Newmeyer, D. D., 165
Niemann, S., 23, 59
Niikura, M., 59, 123
Nishikawa, M., 27
Nisoli, E., 107
Nocca, G., 49, 50
Normoxia, 77
See also Hypoxia
Nuclear DNA mutations and mitochondrial
proteins, 47
Nuclear respiratory factors 1 and 2 (NRF-1 and
NRF-2), 60
Nyengaard, J. R., 113
O
Oakes, S. A., 48
Oberley, L. W., 65
Oberley, T. D., 65
Oca-Cossio, J. A., 67
Odstrcil, E. A., 11
Oesterle, D., 9
Ogunshola, O., 78
Ohh, M., 138–140
Ohta, K., 61
Ohtani, N., 48
Old, L. J., 110
Onaya, T., 61
Opferman, J. T., 48
Oprysko, P. R., 134
Orr, W. C., 96
Orsini, F., 42
Ortega, A. D., 2, 9–11
Osmanian, C., 137
Osthus, R. C., 7
Ostronoff, L. K., 11
Oxidative stress and senescence, 95–96
Oxygen
homeostasis, 73
sensors, 136
Ozben, T., 27
P
Packer, L., 64
Pagano, M., 23
Paik, J. H., 42
Paik, S. G., 106
Palacios, C., 114
Pandolfi, P. P., 91
Pandolfi, S., 140
Pang, S. F., 28
Pan, Y., 134
Index
Panza, C., 168
Papandreou, I., 10
Papa, S., 60
Paragangliomas (PGLs), 23
Parfait, B., 22
Parkes, T. L., 96
Park, J. A., 79
Park, J. H., 172
Park, J. Y., 10
Park, S., 140
Park, S. Y., 11
Pasteur effect, 20
Pasteur, L., 2
Pastorino, J. G., 104
Patel, L., 64
Patino, W. D., 10, 100, 132
Paulin, F. E., 9
Paulus, P., 7
Pawlu, C., 9
Payne, G., 170
Pecina, P., 9, 10
Pecqueur, C., 21
Peczkowska, M., 9
Pedersen, P., 104
Pedersen, P. L., 6–9, 27, 48, 57
Pelicano, H., 10, 113, 126, 171, 172
Penta, J. S., 121, 122, 124
Perez, J. M., 30
Peroxisome-proliferator activated γ coactivator
1 (PGC-1) gene family, 60
Peskin, B. S., 107
Petrelli, J., 107
Petros, J. A., 9, 58, 64, 127
Pette, D., 22
Peutz-Jeghers syndrome, 12
Pfeiffer, K., 20, 58, 59, 62
Pfeiffer, T., 2, 5, 9
Pfister, M. F., 11
PGC-1 related coactivator (PRC), 61
Pheochromocytomas (PHEOs), 23
Phillips, J. P., 96
Phosphoenolpyruvate carboxykinase
(PEPCK), 49
Phosphofructokinase 1 (PFK-1), 84
Phosphoglycerate kinase (PGK), 85
Piana, G. L., 110
Pich, S., 66
PI3K/AKT signaling pathway and VEGF
expression, 78–79
Pilcher, K., 11
Pilkington, G. J., 140
Pineda, G., 110
Pinedo, C. M., 114
Index
Piva, R., 23
Pizarro, A., 168
Pizzo, P., 62
Plas, D. R., 8, 12, 132, 171
Plasminogen-activator inhibitor 1 (PAI-1), 61
Plate, K. H., 135
Platelet-derived growth factor (PDGF)-induced
mitogenic signaling, 35
Podda, A., 23
Polyak, K., 57, 125, 126
Poole, L. B., 35
Porcelli, A. M., 127
Porin-like protein voltage-dependent anion
channel (PPVAC), 48
Posener, K., 5
Pouyss´egur, J., 77, 80, 172
Powis, G., 29
Poyton, R. O., 104, 113
Pozzan, T., 62
Preston, G., 47
Prosser, R., 22
p53 tumor suppressor protein, 8–9
p53 tumor suppressor gene and organismal
life span, 99–100
Pugh, C. W., 140
Puigserver, P., 61
Pyruvate dehydrogenase (PDH), 86
Pyruvate kinase (PK), 86
Q
Quesada, N. M., 38
Quevedo, C., 30
Quintanilla, M., 168
R
Rabbitts, P. H., 92
Racker, E., 9
Raff, M. C., 173
Raghunand, N., 7
Rai, E., 128
Ramanathan, A., 104, 113
Ranganathan, A. C., 41
Ratcliffe, P. J., 8, 136, 140
Rathmell, J. C., 8, 171
Rauscher, F. J., 64
Ravi, R., 137
Reactive oxygen species (ROS), 33
and carcinogenesis, 105
dependent stabilization of HIF-1α, 39
and redox signaling, 34–35
related signaling pathways, 50
ROS-mediated cell growth, molecular
mechanism(s), 64
signal transducing molecule, 34
191
Rechcigl, M. Jr., 66
Redondo A., 9–11
Redox-based signaling, 35
Redox signaling pathway, 63–65
Reed, J., 169
Reed, J. C., 4, 11, 20, 173, 174
Regan, K. J., 38, 40
Reichert, M., 172
Remy, C., 7
Respiration, 3
Respiratory chain (RC)
biochemical and genetic features, 20
cell proliferation and, 25
genetic origin and organization, 21
redox status, 22
and tumors, 29–30
Retrograde signaling in cancer cell
altered antioxidant defense pathway, 65
altered expression of OXPHOS regulatory
genes and subunits, 60–61
altered mitochondrial morphology and,
65–66
antiapoptotic genes activation, 66–67
calcium signaling pathway activation,
62–63
genes activation involved in tumor invasion
and progression, 61–62
mechanism of, 67–68
redox signaling pathway activation, 63–65
Reynier, P., 59
Reynolds, T. Y., 134
Reznick, R. M., 12
Rhee, S. G., 34
Rho GTPase-activating protein, 39
Ricchetti, M., 22
Rice-Evans, C., 63
Rice, K., 137
Richards, F. M., 51
Richard, S. M., 124, 125
Richhardt, N., 11
Ricquier, D., 21
Ridnour, L. A., 65
Riedl, S. J., 166
Rigo, P., 7
Ristow, M., 2, 11
Ritchie, J. M., 29
Rivas, A. L., 114
Roberts-Thomson, S. J., 62
Robey, I. F., 8
Robey, R. B., 171
Rockwell, S., 134
Rodien, P., 59
Rodriguez-Enriquez, S., 46
192
Roeder, L. M., 115, 116
Roe, J. S., 140
Roe, R., 135
Roifman, C. M., 110
Romer, L. H., 169
Rosenberger, S. F., 57, 64, 65
Rosen, O. M., 7
Ross, A. J., 165
Roth, J. C., 137
Roth, P. H., 8, 133, 135, 136
Rotig, A., 22
Rouault, T. A., 4
Roux, D., 77
Rowan, A. J., 25, 59
Roy, D., 50
Rue, E., 135
Rue, E. A., 135
Ruiz, C. R., 114
Ruiz-Pesini, E., 9, 58, 64, 120, 127
Rustin, P., 22–25, 27
Ryan, K. M., 138
Ryu, J. H., 172
S
Saavedra, E., 46
Saccharomyces cerevisiae, 2, 106
arrest defective-1 (ARD1) in, 79
Crabtree effect, 106
Sadlock, J., 22
Sakamaki, T., 11
Sakamoto, K., 12
Sakashita, M., 7
Sala, S., 2, 9–11
Salas, A., 22, 47, 126–128
Salinas, M., 11
Saliou, C., 64
Salvesen, G. S., 166
Samuels, D. C., 124–126, 128
S´anchez-Arag´o, M., 2, 6, 7, 9–11
Sanchez, L. B., 4
Sansone, P., 42, 140
Santamaria, G., 10–12
Santar´en, J. F., 11
Saretzki, G., 96
Saridin, F., 168
Savagner, F., 59–61
Sawyer, D. B., 48
Scarpulla, R. C., 106
Scatena, R., 49–51
Schade, A. L., 5
Schagger, H., 20
Scheid, A., 78
Schimmer, A. D., 167
Index
Schmeller, N., 124
Schmid, T., 138
Schmitt, C. A., 171
Schneider, P., 149
Schon, E. A., 22, 128
Schreiber, R. D., 110
Schreiber, S. L., 104, 113
Schuler, M., 167
Schulz, T. J., 11
Schumacker, P. T., 173
Schuster, S., 2, 5, 9
Scorrano, L., 12, 48, 174
Scott, C. L., 167
Seftor, E. A., 7
Sekido, Y., 139
Sekito, T., 67, 68
Selak, M. A., 9, 24, 25, 47
Semenza, G. L., 8, 10, 75, 81, 92, 132–136,
135, 136, 139
Senescence, 94
bypassing effect of enhanced glycolysis,
96–97
oxidative stress and, 95–96
Senoo-Matsuda, N., 23
Serizawa, S., 11
Serrano, M., 94, 95
Serra, V., 96
Seth, R. B., 110
Shah, A., 7
Sharma, R. I., 9
Sharma, S., 128
Sharma, S. V., 169
Shaw, R. J., 12
Shay, J. W., 94
Sheldon, D. R., 150
Shephard, H. M., 48–50, 58, 61–63, 67
Sherr, C. J., 94
Shidara, Y., 58, 123, 127
Shi, H. H., 57
Shi, J., 35
Shim, H., 7, 93, 171
Shim, Y. J., 107
Shin, Y. K., 10
Shi, X., 64
Shi, Y., 166
Shi, Y. H., 135
Shoshan-Barmatz, V., 12
Shoubridge, E., 128
Shulman, G. I., 12
Simonetti, S., 22
Simoni, D., 150
Simon, M. C., 24, 113, 134
Simonnet, H., 58, 59, 62
Index
Sinclair, D. A., 98
Singh, G., 120
Singh, K. K., 9, 10, 47, 62
Sinkovics, J. G., 150
Sivridis, E., 8
Sledge, G. W. Jr., 64
Sligh, J. E., 8
Smallwood, A. C., 139
Smith, T. A., 9
Sohal, R. S., 96
Sole, P. D., 49, 50
Sorcinelli, M. D., 165
Soslau, G., 66
Spector, M., 9
Spencer, B., 115
Sperl, W., 124
Srivastava, S., 60, 61
Steinbach, J. P., 172
Stengel, P., 75, 77
Stiehl, D. P., 80, 85
Stiles, B., 10
Storci, G., 42, 140
Strasser, A., 137
Streuli, C. H., 169
Strompf, L., 23
Subbaram, S., 37, 38, 40
Succinate dehydrogenase (SDH)
deficiency and cancer, 88
mutations, 23
Sugai, T., 9
Suk, K., 172
Sukosd, F., 57
Sung, H. J., 10
Sun, L., 110
Sun, W., 127, 128
Sun, Y., 61
Supra, P., 172
Swartzendruber, D., 152
Swift, A. L., 9
Swift, H., 66
Symons, M., 168
Sympson, C. J., 168
Synthesis of Cytochrome c Oxidase
2 (SCO2), 10
p 53 gene, 132
T
Tachibana, K., 11
Tait, A. S., 8
Takahashi, A., 48
Takeda, K., 34
Takenaga, K., 172
Tang, D. G., 48
193
Tarassov, I., 22
Taylor, R. W., 9
Tchernyshyov, I., 10
Tedesco, L., 107
Tekaia, F., 22
Tennant, D. A., 25
Terashima, M., 9
Teras, L. R., 107
Testa-Parussini, R., 107
Thierbach, R., 11
Thilly, W. G., 57, 58, 125, 126
Thioredoxin (TRX)
and cysteine, 74–75
overexpression and cell proliferation, 28
Thomas, P. A., 7
Thompson, A. M., 9
Thompson, C. B., 8, 11, 12, 108, 113, 132,
165, 170, 171, 173
Thornton, J., 67, 68
Thun, M. J., 107
Tian, W. N., 97
Tiller, G. E., 8
Timofeev, O., 11
TNF-α/CHX–induced apoptosis, 36
Tolomeo, M., 150
Tomiyama, A., 11
Tomlinson, I. P., 25, 47, 59
Tomlinson, J. P., 121
Tonello, C., 107
Tonks, N. K., 35
Tory, K., 139
Tothova, Z., 42
Tovar, J., 4
TP53-induced glycolysis and apoptosis
regulator (TIGAR), 9, 100
Traber, M. G., 64
Transcription factor Ets-1, 38
Trauger, R., 152
Tremble, P., 168
Troncoso, P., 175
Tsao, T., 174
Tschopp, J., 149
Tsuda, M., 23
Tsuruo, T., 149
Tsuruta, A., 9
Tu, B. P., 11
Tumor
cell mitochondria, 66
glucose-specific transporters (GLUTs)
induction, 7
metastasis and, 39
microenvironment for, 136–137
PTEN suppressor, 40
194
Tumor (cont.)
superoxides role in, 27
tumorigenesis
temporal and spatial effects in, 112–113
tumor-specific metabolic pathways,
146–147
VEGF role, 38
Turnbull, D. M., 9, 124, 125
Tyner, S. D., 100
Tyurina, Y. Y., 107
Tyurin, V. A., 107
Tzagoloff, A., 19, 21
U
Uesugi, N., 9
UnCoupling proteins (UCP), 21–22
uncoupling protein 2 (UCP2) expression in
colon cancers, 110
Usadel, H., 57
Usher, P., 7
V
Vahsen, N., 12, 20, 166
Vainio, H., 107
van Beest, M., 140
Vander Heiden, M. G., 11, 108, 171, 173
Vande Velde, C., 172
Vandromme. M., 135
van Holde, K. E., 106, 113
van Waveren, C., 61
Vega, A., 22, 47, 126–128
Velankar, M., 174
Velours, J., 11
Venditti, J. M., 150
Venkatachalam, S., 100
Verma, A., 8, 82, 136
Vijayasarathy, C., 48–50, 58, 61–63, 67, 119
Vijayvergiya, C., 64
Villalobos-Menuey, E., 152
Villalobos-Menuey, E. M., 155
Villani, G., 60
Villunger, A., 137
Vincent, B. J., 8
Vincenzoni, F., 50
Voest, E. E., 140
Voet, D., 146, 147
Voet, J., 146, 147
Vogelstein, B., 137, 169
Voigt, A., 11
Volmat, V., 77
von Hippel–Lindau syndrome, 8, 138–140
VHL gene, mutation in, 81
von Hippel–Lindau tumor suppressor
protein (pVHL), 75
Index
von Kienlin, M., 7
von Zglinicki, T., 96
Vousden, K. H., 138
W
Wachsman, J. T., 121, 122, 124
Waddle, J. A., 68
Wahli, W., 51
Walensky, L. D., 165
Wallace, D. C., 25, 27, 47, 119–122,
120, 127, 128
Walsh, S. A., 7
Wan, X. S., 65
Wang, C., 11, 104, 110, 113
Wang, G. L., 8, 133, 135, 136
Wang, J., 94
Wang, S., 10
Wang, T., 11
Wang, X., 4
Wang, Y., 94
Wang, Z. Q., 113
Warburg, O., 5, 6, 20, 56, 81, 120, 126,
131–133
Warburg hypothesis
hypoxia inducible factor 1 alpha (HIF-1α)
and, 8
model, 113–115
principles of, 7
Warren, L., 66
Watson, A., 51
Watson, P. H., 137
Watt, F. M., 169
Waugh, T. A., 29
Way, I. P., 169
Webster, K. A., 114, 173
Weiler, S., 165
Weinberg, R. A., 11, 92, 134
Weinhouse, S., 5
Weller, M., 172
Welsh, S. J., 8, 29
Wenger, R. H., 78, 80, 85, 138
Weng, Y., 139
Wen, S., 175
Werb, Z., 168
Werner, E., 168
Weydert, C. J., 29
Whang, E. E., 169
Whelan, M., 149
Wiesener, M. S., 136
Wilhelm, M., 57
Wilkie, D., 66
Willett-Brozick, J. E., 9, 20, 23, 59
William, O., 108
Index
Williamson, J. R., 113
Wind, F., 20
Wolff, T., 9
Wong, G. H. W., 36
Woods, M., 5
Wood, Z. A., 35
Worley, J. F., 115
Wright, W. E., 94
Wu, H., 10
Wu, L., 137
Wu, M., 9
Wu, X., 91
Wu, Z., 61
Wurster, R. D., 63
Wykoff, C. C., 137, 140
X
Xia, Y., 57
Xie, L. Y., 94
Xue, W., 171
Xue, Y., 65
Xu, J. N., 28
Xu, Q., 11, 140
Xu, R. H., 10, 113, 171, 172
Xu, R. K., 28
Y
Yamagata, K., 58, 123, 127
Yamakoshi, K., 48
Yamauchi, A., 125
Yang, H., 139
Yang, M., 171
Yang, Q. H., 28
Yao, Y. G., 22, 47, 126–128
Yasuda, K., 23
Yauch, R. L., 140
YC-1–treated Hep3B hepatoma cells, aldolase
and enolase mRNAs, 85
Yeast, Sir2 as longevity gene in, 97–98
calorie restriction and, 98–99
195
Yee, A. J., 11
Yen, P., 29
Yoo, B. C., 10
Yoon, B. I., 29
Younes, M., 7
Youn, H. D., 140
Yuan, F., 134
Yu, Z. X., 34
Z
Zaidi, M., 63
Zamzami, N., 30
Zandi, E., 64
Zanssen, S., 128
Zatyka, M., 140
Zbinden, I., 64
Zeamari, S., 140
Zeiher, A. M., 170
Zeitlin, B. D., 174
Zeller, K. I., 94
Zender, L., 171
Zhang, S. Y., 58, 61, 62, 119
Zhao, Y., 65
Zheng, J. Z., 135
Zhong, H., 80
Zhou, J., 79, 138
Zhou, S., 10, 127, 128
Zhou, X. M., 170
Zhou, Y., 172
Zhu, H., 57, 125, 126
Zhu, T., 12
Ziegler, A., 7
Zigmond, M. J., 41
Zinner, M. J., 169
Zong, W. X., 113, 165
Zou, Y., 96
Zummo, G., 48
Zu, X. L., 5
Zweier, J. L., 57