Download HYPOXIA AND THE METABOLIC PHENOTYPE OF PROSTATE CANCER CELLS

HYPOXIA AND THE METABOLIC PHENOTYPE OF
PROSTATE CANCER CELLS
by
Lauren Higgins
A thesis submitted to the Department of Biology
In conformity with the requirements for
the degree of Master of Science
Queen’s University
Kingston, Ontario, Canada
September, 2008
Copyright © Lauren Higgins, 2008
Abstract
Cancer cells have the ability to survive when oxygen is limiting, and upregulate
the pathway of fatty acid synthesis, owing in part to alterations in their metabolism. I
compared the metabolic phenotypes of the prostate cancer cell lines LNCaP, DU145, and
PC3 assessing energy metabolism, and metabolic gene expression. I also explored the
plasticity of the metabolic phenotype following passage, selection and in vivo growth.
Finally, I explored the sensitivity of the fatty acid synthesis pathway to low oxygen.
LNCaP cells had a more oxidative phenotype based on oxygen consumption,
lactate production, enzyme assays, and mRNA levels. While DU145 and PC3 cells were
more glycolytic, they were unresponsive to dichloroacetate (DCA), and dinitrophenol
(DNP), stimulators of oxygen consumption. Mitochondrial dysfunction in the PC3 and
DU145 cells may explain this phenomenon, for they possessed normal cardiolipin levels
but lower mitochondrial enzyme activities (cytochrome oxidase (COX), citrate synthase
(CS)).
When LNCaP cells were subjected to high passage, with and without clonal
selection, the derived lines acquired a dysfunctional oxidative phenotype, becoming more
glycolytic. Clonal selection appeared to have the most dramatic effect on cellular
metabolism. This finding is supported by decreased oxygen consumption, increased
lactate production, and a reduction in the activity of the oxidative enzymes CS and COX
in the clonally selected LNCaP-luc cell line. Similar to the DU145 and PC3 cells, NAO
fluorescence indicates that the oxidative impairment in these LNCaP-derived lines may
be due to a reduction in mitochondrial activity. The pattern of metabolic gene expression
ii
seen in vitro was unaffected when LNCaP cells were grown as subcaspular and muscle
xenografts in immunodeficient mice, though xenografts did exhibit indications of an
hypoxic response (elevated VEGF mRNA).
Oxygen deprivation in vitro increased mRNA for HIF and responsive genes but
not SREBP responsive genes. Similarly, oxygen deprivation had no influence on
triglyceride levels in any of the lines suggesting that the SREBP axis may not be directly
modulated by oxygen levels.
Collectively these studies demonstrate differences in the metabolism of these
prostate cancer models, with important ramifications of therapeutic strategies involving
metabolic targets.
iii
Co-Authorship
Chapter 2, “Hypoxia and the metabolic phenotype of prostate cancer cells,” was
co-authored with A. Garbens, H. Withers, and Dr. C.D. Moyes from Queen’s University
and Dr. S. Hayward and H. Love from Vanderbilt University. Dr. Moyes was the
supervisor for this work, contributing to experimental design, troubleshooting and
editing.
Alaina Garbens performed selected real-time PCR experiments to evaluate the
sensitivity of the SREBP axis to low oxygen as a fourth year thesis project. I completed
and extended this work after her graduation.
Henry Withers, a former undergraduate student in our lab, did enzyme assays to
compare the levels of metabolic enzymes between selected cell lines. I completed and
extended this work after his graduation.
Dr. Hayward and Harold Love from Vanderbilt were responsible for sending us
the xenografts in SCID mice. They performed the injections, and collected the
xenografts. They also provided us with aliquots of highly passaged and luciferase
transfected cells, although the cells and experiments were run here.
Leo Magnoni from the University of Ottawa (supervisor JM Weber) was kind
enough to demonstrate to me how to perform my lipid extractions in their lab.
iv
Acknowledgements
Many thanks to Chris Moyes (The Big Kahuna, Boss, Wizard, Professor M, P
Moyes) for being the best supervisor, mentor, and friend I could have hoped for.
I would also like to thank my fellow members of the Moyes lab, for I would not
have made it through without them! Thanks to Chris Le Moine for teaching me how to
keep my cool in every aspect of my life and for welcoming me warmly into the Moyes
lab. Thank you to Melanie Fortner, for teaching me (basically) everything I know in the
lab. Thank you to Rhiannon Davies for being an amazing friend and for keeping me
entertained in Kingston. Thank you to Alex Little for making me laugh. I would also like
to thank Christine Genge for her friendship and moral support. Thanks to Henry Withers
and Alaina Garbens for learning alongside me, and for all of their help with my project.
I give sincere thanks to my committee members Dr. Chris Mueller and Dr.
Virginia Walker for all of their patience and guidance throughout my thesis work. Thank
you to Bill Bendena, Steve Lougheed, and Paul Young for being invaluable mentors to
me. I would also like to thank Dr. Weber at the University of Ottawa for allowing me to
work in his lab to do my lipid extractions. Many thanks to the Chew lab for all of their
advice and for helping me prepare for my defense.
Finally, I would like to thank my friends and family for giving my life balance
these past two years. Amy Hewitt, Ted Branscombe, Danae Benjamin, Emily Huva, Gary
Armstrong, and Anthea Christophorou – you have been there for me throughout some of
the best and worst times of my life and I will love you forever for that. I would like to
thank my grandparents, Meryl and Don Forrest for being my academic backbones and
v
supporting me tremendously throughout my academic career. I would lastly like to thank
my mother, Jane, and father, Tim, for all of their unconditional love and support.
vi
Table of Contents
Abstract................................................................................................................... ii
Co-Authorship ....................................................................................................... iv
Acknowledgements .................................................................................................v
Table of Contents.................................................................................................. vii
List of Tables ...........................................................................................................x
List of Figures........................................................................................................ xi
List of Abbreviations ........................................................................................... xiii
Chapter 1: Introduction and General Literature Review .........................................1
1.1. Overview ...............................................................................................................1
1.2. Cellular Metabolism ..............................................................................................2
1.3 Aerobic glycolysis in cancer...................................................................................4
1.3.1 Hypoxic selection ....................................................................................4
1.3.2 Mitochondrial dysfunction ......................................................................5
1.3.2 Activation of Akt/PKB............................................................................7
1.3.3 Inactivation of p53 ................................................................................10
1.4 Hypoxia and cellular metabolism .........................................................................12
1.4.1 Hypoxia inducible factor-1 (HIF-1) transcription factor.......................13
1.4.2 Angiogenesis .........................................................................................13
1.4.3 Glucose uptake and glycolysis ..............................................................15
1.4.4 Apoptotic resistance ..............................................................................16
1.5 Fatty acid synthesis and cancer ............................................................................19
vii
1.5.1 Sterol regulatory element-binding protein (SREBP) transcription factors
....................................................................................................................................21
1.5.2 Androgens and lipogenesis....................................................................23
1.5.3 Hypoxia and fatty acid synthesis...........................................................24
1.6 Objectives .............................................................................................................27
Chapter 2: Hypoxia and the metabolic phenotype of prostate cancer cells...........30
2.1 Introduction ..........................................................................................................30
2.2 Methods ................................................................................................................32
2.2.1 Cell culture and treatments....................................................................32
2.2.2 Oxygen consumption and lactate production measurements ................34
2.2.3 Enzyme assays.......................................................................................35
2.2.4 RNA extraction and real-time PCR.......................................................36
2.2.5 Intracellular ATP levels ........................................................................38
2.2.6 Cardiolipin content (NAO staining) ......................................................38
2.2.7 Xenografts in SCID mice ......................................................................39
2.2.8 Triglyceride extraction and quantification ............................................39
2.2.9 Statistical analyses.................................................................................40
2.3 Results ..................................................................................................................40
2.3.1 The metabolic phenotype differs between cell lines .............................40
2.3.2 DU145 and PC3 cells have reduced mitochondrial enzyme activity but
not mitochondrial gene expression.............................................................................44
viii
2.3.3 LNCaP cells transition to a glycolytic phenotype with increasing
passage and clonal selection.......................................................................................49
2.3.4 Metabolic gene expression of LNCaP cells in vivo...............................53
2.3.5 SREBP axis gene expression is not hypoxia sensitive ..........................57
2.3.6 Fatty acid levels are not influenced by low oxygen in these prostate
cancer models .............................................................................................................59
2.4 Discussion.............................................................................................................59
2.4.1 Metabolic enzyme activities and gene expression ................................60
2.4.2 The effects of passage and selection .....................................................63
2.4.3 In vivo and in vitro differences in gene expression ...............................64
2.4.4 SREBP axis ...........................................................................................66
Chapter 3: General Discussion ..............................................................................68
3.1 The metabolic phenotype of prostate cancer cells................................................68
3.2 The influence of passage and selection on prostate cancer cells and xenografts .71
3.3 The influence of hypoxia on fatty acid synthesis .................................................74
Appendix 1 ............................................................................................................91
ix
List of Tables
Table 1. Real-time PCR human specific primers (5’-3’)...................................................37
Supplementary Table 1. Raw cycle thresholds of real-Time PCR data comparing LNCaP,
DU145, and PC3 cells (Figure 7) ..............................................................................92
Supplementary Table 2. Raw cycle thresholds of real-Time PCR data comparing LNCaP
and derived lines (Figure 11)………………………………………..………………..….92
Supplementary Table 3. Raw cycle thresholds of real-Time PCR data comparing HPLNCaPs and xenografts (Figure 12)…………………………………………..…………93
Supplementary Table 4. Raw cycle thresholds of real-Time PCR data comparing LNCaPluc cells and xenografts (Figure 13)…………………………………………...…………94
Supplementary Table 6. Raw cycle thresholds of real-Time PCR data comparing LNCaP,
DU145, and PC3 cells following low oxygen treatment (Figure 14)…………………....95
x
List of Figures
Figure 1. Possible ways that low oxygen may influence the SREBP axis via HIF-1α .....22
Figure 2. Metabolism of LNCaP, PC3, and DU145 cells..................................................41
Figure 3. Lactate production..............................................................................................42
Figure 4. ATP levels following exposure to low pH and hypoxia ....................................43
Figure 5. Metabolism of LNCaP, DU145, and PC3 cells following treatment with
dichloroacetate (DCA)...............................................................................................44
Figure 6. Glycolytic and mitochondrial enzyme activities in DU145 and PC3 cells ........45
Figure 7. mRNA levels in LNCaP, DU145, and PC3 cells ...............................................47
Figure 8. Oxygen consumption of LNCaP, DU145 and PC3 cells following exposure to
dinitrophenol (DNP) ..................................................................................................48
Figure 9. NAO fluorescence in LNCaP, DU145, PC3, and LNCaP-derived cell lines
measured using FACS ...............................................................................................49
Figure 10. Metabolism of HP-LNCaP and LNCaP-luc cells.............................................51
Figure 11. Transcript levels in HP-LNCaP and LNCaP-luc cells .....................................52
Figure 12. Transcript levels in HP-LNCaP cells and xenografts.......................................54
Figure 13. Transcript levels in LNCaP-luc cells and xenografts.......................................56
Figure 14. mRNA levels
in LNCaP, DU145 and PC3 cells following exposure to
hypoxia, anoxia, or azide ...........................................................................................58
Figure 15. Triglyceride content (ug TG/mg protein) of LNCaP, DU145, and PC3 cells
following exposure to normoxia, hypoxia, and anoxia. ............................................59
xi
Supplementary Figure 1. Integration of Cellular Pathways that Mediate Aerobic
Glycolysis. .................................................................................................................91
xii
List of Abbreviations
ALDO: aldolase; CS: citrate synthase; COX: cytochrome c oxidase; DCA:
dichloroacetate; Δψm: mitochondrial membrane potential; G3PDH: glyceraldehyde-3phosphate
dehydrogenase;
G6PDH:
glucose-6-phosphate
dehydrogenase;
HK:
hexokinase; HIF: hypoxia-inducible factor; LDH: lactate dehydrogenase; NRF-2; nuclear
respiratory factor 2; OXPHOS: oxidative phosphorylation; PBS: phosphate-buffered
saline; PDH: pyruvate dehydrogenase; PDK: pyruvate dehydrogenase kinase; PGC-1:
PPARγ co-activator 1; PGI: phosphoglucoisomerase; PK: pyruvate kinase; PKm:
pyruvate kinase muscle isoform; SCAP: SREBP cleavage activator protein; SCID: severe
combined immunodeficiency; SCO2: synthesis of cytochrome c oxidase 2; SREBP:
Sterol regulatory-element binding protein; TBP: TATA-binding protein; VEGF: vascular
endothelial growth factor.
xiii
Chapter 1: Introduction and General Literature Review
1.1. Overview
The metabolic phenotype of cancer cells is often neglected in cancer research,
which typically focuses on mutational mechanisms of cancer progression. Nonetheless,
some efforts have been placed in understanding cancer cell physiology, specifically the
ability of cancer cells to survive and progress when oxygen is limiting. The focus of this
study is to characterize the metabolic properties of the three most widely used prostate
cancer cell lines (LNCaP, DU145, and PC3). We will assess the profile of metabolic
enzymes in terms of catalytic activity and expression of the genes encoding the enzymes
and their transcriptional regulators. We will also evaluate how the metabolic phenotype
changes under various culture conditions (prolonged passage, clonal selection, and
xenograft models). Lastly, we will explore the relationship between the sterol regulatory
element binding protein (SREBP) pathway and low oxygen, to determine if it is directly
influenced by low oxygen in prostate cancer cells. The goal is to determine the
relationships between hypoxia, mitochondrial function, fatty acid synthesis and the
phenotype of aerobic glycolysis. A detailed understanding of prostate cancer metabolism
at the molecular level is crucial to understanding the progression of this disease, and may
lead to the identification of possible therapeutic targets.
1
1.2. Cellular Metabolism
In order to produce ATP, cells break down carbohydrates, proteins and fatty
acids, and use the resulting energy to support anabolic pathways necessary for cellular
growth and replication, such as DNA replication and protein synthesis (Costello and
Franklin 2005). Energy metabolism, in the context of my thesis, refers to the combined
processes of glycolysis, the tricarboxylic acid cycle (TCA cycle) and oxidative
phosphorylation. It is tightly regulated according to cellular energy demands.
Glycolysis, the first major pathway of cellular metabolism, occurs in the
cytoplasm and is functional even in the absence of oxygen. While glycolysis is able to
produce ATP at a high rate, it is considered a low efficiency pathway because it produces
a net of only two ATP per glucose. A summarized equation of glycolysis is as in
Equation 1.1:
Eq 1.1. C6H12O6 + 2ADP + 2NAD+  2ATP + 2 pyruvate + 2NADH +2H+
The pyruvate that is produced in the cytoplasm is imported into the mitochondrial
matrix, where it is converted by pyruvate dehydrogenase (PDH) into acetyl-coenzyme A
(acetyl-CoA). These molecules of acetyl-CoA then enter the TCA cycle where they are
further oxidized to produce 2 CO2, 3 NADH and 1 FADH2. The reducing equivalents
(NADH, FADH2) donate their electrons to the electron transport chain, and during the
electron transfers within and between complexes, protons are pumped into the
intermembrane space. This generates an electrochemical gradient (the proton motive
2
force), which can be harnessed by ATP synthase to form ATP from ADP and an
inorganic phosphate. Historically, reducing equivalent oxidation was thought to pump
enough protons to create a proton motive force sufficient to produce 3 ATP from NADH
and 2 ATP from FADH2. As well, the reducing energy produced in glycolysis (2 NADH
per glucose) can be shuttled into the mitochondria, where it can be oxidized to produce 2
or 3 ATP (per glucose). Thus, oxidative phosphorylation can produce as many as 38 or
39 ATP per glucose, in contrast to 2 ATP per glucose in glycolysis. Though the exact
stoichiometries for ATP production per NADH and FADH2 are probably lower than 3
and 2, by any measure oxidative phosphorylation is much more efficient in terms of ATP
per glucose than is glycolysis.
Another difference in the two pathways is the dependence on oxygen. Oxidative
phosphorylation requires oxygen as a final electron acceptor, and thus in the absence of
oxygen, mitochondria can neither pump protons nor generate ATP. Under in the absence
of oxygen, cell must rely on the ATP derived from glycolysis for survival. Since NADH
produced in glycolysis cannot be oxidized via mitochondria redox shuttles, cells balance
redox (NADH/NAD+) by converting pyruvate into lactate by the enzyme lactate
dehydrogenase (LDH). As pyruvate is reduced, forming lactate, NADH produced during
glycolysis is reoxidised, forming NAD+. Glycolysis is essential during periods of hypoxia
and anoxia but its inherent inefficiency (in terms of ATP per glucose) is a problem for
otherwise normoxic, healthy cells. Thus, it would be unusual for normal cells to rely on
glycolysis under aerobic conditions (termed aerobic glycolysis).
3
1.3 Aerobic glycolysis in cancer
In contrast to healthy cells, many types of cancer cells prefer glycolysis as an
ATP source, even in the presence of abundant oxygen (Warburg 1926; Pedersen 1978).
Why a cell would settle for two molecules ATP per molecule of glucose rather than the
additional 34 molecules of ATP from aerobic pathways remains hotly debated. Numerous
explanations have been put forward including hypoxic selection, mitochondrial
dysfunction, activation of Akt, and deactivation of the tumour suppressor p53. I will
consider each of these mechanisms in turn, however, none can be generalized to all cell
types or types of cancer and a combination of mechanisms are likely involved.
1.3.1 Hypoxic selection
Somatic evolution during carcinogenesis involves changes in the activity of
tumour suppressor genes and proto-oncogenes. Such mutations confer a selective
advantage to the cell, and contribute to malignant transformation (Gillies and Gatenby
2007). Molecular changes during the evolution of carcinogenesis are driven by selective
pressure within the cellular microenvironment (Gillies and Gatenby 2007). Early in
tumorigenesis, prior to vascularization, the cells in the centre of the tumour are unable to
access oxygen and nutrients from the bloodstream and experience periods of intermittent
hypoxia. Under such circumstances, enhanced glycolysis may benefit the cell in two
ways: 1) by maintaining ATP production when oxygen is limiting, and 2) by producing
reducing equivalents that protect the cell from oxidative stress during periods of reoxygenation (Gillies and Gatenby 2007). However, cancer cells maintain their glycolytic
4
phenotype following vascularization. Therefore, selection for cancer genotypes that
improve glycolysis under anaerobic conditions may result in a genotype that must rely on
glycolysis even under aerobic conditions.
1.3.2 Mitochondrial dysfunction
Warburg’s original explanation for aerobic glycolysis in cancer cells was
irreversible damage to the respiratory apparatus (Warburg 1926). Mitochondria are
arguably the most important component of the cell for they are responsible for the
production of 95% of the average eukaryotic cell’s energy and play a central role in
metabolism (Goffart and Wiesner 2003). Consequently, many lines of evidence suggest
that the underlying cause of the aerobic glycolytic phenotype in cancer cells may be due
to mitochondrial shortfalls or dysfunction (Warburg 1930; Pedersen 1978). It is known
that mitochondrial DNA (mtDNA) mutations and deletions can occur in some contexts
where aerobic glycolysis is observed (Wallace 1999; Xu et al. 2005). Alternatively, it is
perhaps more likely that the primary metabolic lesion is a dysregulation of mitochondrial
biogenesis: cancer cells may derive the majority of their ATP from the glycolytic
pathway because they do not have a sufficient number of otherwise normal mitochondria
to yield ATP from oxidative metabolism. When the energy demands of a cell are
heightened, a transcriptional network is activated to increase the content of mitochondria
in the cell. This is most notable in exercising muscle (Hood 2001) and during adaptive
thermogenesis in brown adipose tissue (Klingenspor et al. 1996), both of which require
an increase in mitochondrial content to maintain the cell’s high demand for energy.
5
Likewise, a reduction in the content of mitochondria per cell may result in a cell that
cannot sustain oxidative ATP production. Many cancer cells have fewer mitochondria per
cell in comparison to healthy cells (Pedersen 1978; reviewed by Moreno-Sanchez et al.
2007). A reduction in mitochondrial content may result from either more dynamic
mechanisms of organelle degradation (Rodriguez-Enriquez et al. 2006) or suppressed
organelle biogenesis. While these observations are widespread, their molecular
underpinnings have yet to be discovered.
Mitochondrial biogenesis and oxidative metabolism are controlled by a number of
genes, most of which are nuclear encoded and are regulated at a transcriptional level
(Scarpulla 1997; Lenka et al. 1998). This regulatory network involves transcription
factors including nuclear respiratory factors (NRF-1, NRF-2), peroxisome proliferatoracivated receptors (PPAR-α, PPAR-γ) and coactivators such as PPAR- γ coactivator 1
alpha family members PGC-1α and PGC-1β. PGC-1α binds to nuclear hormone receptors
and other transcription factors to induce the transcription of their target genes involved in
fuel intake, fatty-acid oxidation, mitochondrial biogenesis, and oxygen consumption (Wu
et al. 1999). This coactivator is highly expressed in brown fat where it is involved in
thermogenesis (Puigserver et al. 1998), muscle where it is implicated in fibre type
switching (Lin et al. 2002), and liver, where it is important for gluconeogenesis (Yoon et
al. 2001). PPAR-γ and PGC-1α mRNA levels are reduced in breast (Jiang et al. 2003),
ovarian (Zhang et al. 2007), and colon cancer cells (Feilchenfeldt et al. 2004).
Furthermore, Zhang et al. (2007) found that PGC-1α induces apoptosis in ovarian cancer
6
cells, suggesting that its low expression in cancerous tissues may be associated with
cancer growth and progression. Therefore, altered expression of these transcriptional
regulators may be involved in the impaired oxidative metabolism observed in many
cancer types.
1.3.2 Activation of Akt/PKB
The phosphatidylinositol 3-kinase (PI3K) signaling pathway, which mediates
many aspects of cell growth and survival, is commonly deregulated in cancer (Bellacosa
et al. 1995; reviewed by Cantley 2002). Activation of this pathway ultimately leads to
growth and metastasis, and has been associated with therapeutic resistance (Bacus et al.
2002; Clark et al. 2002; reviewed by Hennessy et al. 2005). PI3K belongs to a family of
serine/threonine kinases that are responsible for internalizing the effects of growth factors
and receptor tyrosine kinases at the cellular membrane leading to the activation of an
internal signaling cascade. PI3K, upon activation, subsequentially phosphorylates
phosphatidylinositol-4, 5-bishosphate resulting in the formation of a secondary
messenger, phosphoinositol-3’phosphate (PIP3). The tumour suppressor, phosphatase and
tensin homologue (PTEN), has the ability to remove this phosphate and reduce levels of
PIP3, thereby interfering with the effects of PI3K (Stambolic et al. 1998). In the absence
of PTEN, PIP3 binds to lipid domains, namely pleckstrin-homology (PH) domains in its
targets to recruit them for activation at the membrane. The principal moderator of the
effects of PI3K activation is the downstream Akt (also known as protein kinase B). Akt is
commonly constitutively expressed in cancer cells, and is very important in the
7
development of cancer for it both mediates aerobic glycolysis and is involved in the
suppression of apoptosis.
In addition to hypoxic selection and mitochondrial dysfunction, activation of Akt
may explain the phenotype of aerobic glycolysis. It has been shown that activation of Akt
leads to increased glucose consumption and lactate secretion by the cell, and thereby
activates glycolysis without effecting cellular respiration (Elstrom et al. 2004).
Furthermore, studies have indicated that impaired respiration my also activate Akt
(Pelicano et al. 2006). Mitochondrial mutagenesis, hypoxia, and pharmacological
inhibition of the respiratory chain have been associated with the inactivation of PTEN,
and the subsequent activation of Akt (Pelicano et al. 2006). Pelicano and colleagues
(2006) demonstrated that this occurs due to the NADH-mediated inactivation of PTEN.
Hence, upon inhibition of mitochondrial respiration, the reducing equivalent NADH
produced in glycolysis is unable to be oxidized via mitochondrial pathways. This increase
in NADH disturbs cellular redox balance, which has been associated with PTEN
inactivation (Pelicano et al. 2006). Upon activation resulting from loss of PTEN activity,
Akt induces glucose transport into the cell, hexokinase (HK) activity and glycolysis as a
means to derive energy when mitochondrial pathways are not functioning. This
ultimately gives malignant cells an energetic advantage when mitochondrial pathways are
injured, and is associated with resistance to chemotherapy. Therefore, Akt is an important
player in the metabolic phenotype of cancer cells and may represent the mediator
8
between mitochondrial dysfunction and the Warburg effect (i.e. accelerated rates of
glycolysis).
There is a strong relationship between growth factor signaling and cellular
metabolism, as both are absolutely necessary to ensure cell survival. This relationship is
mediated by Akt, which not only promotes aerobic glycolysis, but is also involved in the
evasion of apoptosis. Early in the apoptotic cascade, cytochrome c and other proapoptotic
proteins are released from the mitochondria. Kennedy et al. (1999) demonstrated that Akt
actually inhibits the release of cytochrome c from the mitochondria. Therefore, activation
of Akt prevents the apoptotic cascade from being initiated. Akt exerts its inhibitory
influence on the apoptotic cascade by increasing the activity of mitochondrial-associated
hexokinases (mtHKs) (Robey and Hay 2005). Hexokinase (HK) is the first enzyme of
glycolysis that is responsible for the conversion of glucose into glucose-6-phosphate in
the cytoplasm. The most prominent isoforms of HK are HKI and HKII, and both
associate with and bind to the voltage dependent anion channel (VDAC) in the outer
mitochondrial
membrane.
This
serves
to
couple
glycolysis
and
oxidative
phosphorylation, and provides mitochondrial-derived energy for HK activity. The
activation of Akt localizes HK to the mitochondria, where it is able to regulate the
opening of the permeability transition (PT) pore (Beutner et al. 1998). Therefore, by
increasing the amount of mtHKs that interact with the VDAC, Akt prevents the PT pore
from opening, thereby preventing the release of apoptotic proteins (Gottlob et al. 2001).
9
The kinase Akt has been the target of many therapeutic studies for its ability to link
cellular metabolism to survival.
1.3.3 Inactivation of p53
Another mechanism by which cancer cells may attenuate their glycolytic
phenotype is through the simple inactivation of p53. The tumour suppressor gene, p53, is
often mutated in cancer cells preventing it from initiating the apoptotic cascade in
response to DNA damage (Hollstein et al. 1999). Hence, p53 controls DNA repair, cell
cycle progression, apoptosis, and senescence in healthy cells (Green and Chipuk 2006).
The high frequency of both p53 mutations and aerobic glycolysis in cancer cells suggests
that these two hallmarks of malignancy may be related. Two candidate proteins have
been identified that may be responsible for linking p53 loss of function with the
phenotype of aerobic glycolysis: the Synthesis of Cytochrome c Oxidase (SCO2) protein
influences respiration and the Tp53 Induced Glycolysis and Apoptosis Regulator
(TIGAR) exerts its effects on glycolysis.
Matoba et al. (2006) discovered that the Synthesis of Cytochrome Oxidase
(SCO2) gene might be the link between aerobic glycolysis and p53. SCO2 is necessary
for formation of the COX holoenzyme in the mitochondria, which is required for proper
functioning of the respiratory chain. Reduced expression of the SCO2 protein therefore
ultimately leads to impaired respiration. The fact that SCO2 expression is p53-dependent,
suggests that the loss of p53 function may be the critical factor responsible for glycolytic
stimulation. SCO2 protein levels and hence respiratory rate have been shown to correlate
10
with p53 gene dosage. Furthermore, the resulting impaired mitochondrial respiratory
chain may also stimulate Akt, which would lead to further glycolytic stimulation (Young
and Anderson 2008).
In a similar manner, Bensaad and colleagues (2006) identified the protein TIGAR
to be expressed according to p53 gene dosage. This p53-inducible protein forces glucose
through the pentose phosphate shunt, thereby negatively regulating glycolysis. By
decreasing levels of fructose-2, 6-bisphosphate, TIGAR blocks glycolysis and leads to
the generation of NADPH by the pentose phosphate way. The purpose of this shunt may
be to direct glucose away from the production of energy so that it can be used to build
molecules required for the synthesis of nucleic acids and other key components required
to repair cell damage and ensure cell survival. This NADPH stimulates glutathione
(GSH), an antioxidant that is involved in scavenging reactive oxygen species (ROS). In
cells that are p53-deficient, TIGAR cannot exert its inhibitory effects on glycolysis.
Under such circumstances, glycolysis is stimulated and the levels of ROS will increase in
the cell due to the absence of GSH. As a result, the increase in cellular ROS may
stimulate Akt by inhibiting PTEN, therefore further stimulating glycolysis.
The paradox by which malignant cells derive most of their cellular energy from
glycolysis rather than the more efficient pathway of mitochondrial respiration can be
explained by a multitude of factors. Hypoxia selection, mitochondrial dysfunction, Akt
activation and loss of p53 function may all be responsible for the altered metabolism of
cancer cells. Periods of intermittent hypoxia in the tumour microenvironment may select
11
for cells that are more glycolytic that are able to survive when oxygen is not available.
Defects in the mitochondrial respiratory chain, from mtDNA mutations or problems with
organelle biogenesis, may also explain why cancer cells would prefer glycolysis as their
primary energy source. Similarly, mutations in oncogenes and tumour suppressor genes
may be responsible for this peculiar phenotype. For example, the simple activation of Akt
or inactivation of p53 can explain the Warburg effect. While all of these represent
explanations for aerobic glycolysis (as observed in Supplementary Figure 1), the precise
source of this phenomenon remains unclear. A combination of mechanisms is likely
responsible for the metabolic phenotype of cancer cells, however, the sequence by which
the phenotype is selected for has yet to be elucidated.
1.4 Hypoxia and cellular metabolism
In the previous section, I considered the possible reason why cancer cells appear
committed to glycolysis, regardless of oxygen levels. However, hypoxia is an important
element of tumour biology. Cancer cells in tumours proliferate rapidly and can outgrow
the existing vasculature, leading to regional hypoxia. As discussed previously, this
hypoxic exposure may lead to the selection of a glycolytic phenotype. Additionally, it
triggers many metabolic and cellular responses in the tumour, including angiogenesis,
glycolytic gene induction, and apoptotic resistance, all which are mediated by the
hypoxia inducible factor-1 (HIF-1) transcription factor.
12
1.4.1 Hypoxia inducible factor-1 (HIF-1) transcription factor
The hypoxia inducible factor-1 (HIF-1) is a heterodimeric transcription factor that
functions as the master regulator of oxygen homeostasis. It is composed of an oxygen
sensitive HIF-1α subunit and a constitutively expressed HIF-1β subunit. Under normoxic
conditions, human HIF-1α is hydroxylated on two proline residues (402 and 564)
allowing for the binding of the Von Hippel-Lindau (VHL) protein, which ubiquinates
HIF-1α thereby targeting it for proteasomal degradation (Epstein et al., 2001; Ivan et al.,
2001). Under hypoxic conditions, the dimerization of the HIF-1 subunits allows for their
interaction with the co-activator p300 (CBP) to induce the expression of a variety of
genes involved in oxygen delivery (erythropoiesis and angiogenesis) and survival during
hypoxia (glucose uptake and glycolysis), all of which contain hypoxia response elements
(HREs) in their promoters (Semenza 2007).
1.4.2 Angiogenesis
Angiogenesis, the process of developing new blood vessels from the existing
vasculature, is necessary to ensure all cells of a multicellular organism have a constant
supply of oxygen (Hoeben et al. 2004). This process of blood vessel formation involves
the replication of endothelial cells, partial degradation of the basement membrane and
surrounding extracellular matrix, endothelial cell migration and formation of a tubular
structure (Hoeben et al. 2004). This process is especially important in tumours, for
without their own blood supply, they have limited access to nutrients and oxygen from
the circulation and are termed avascular. Once this ‘angiogenic switch’ has occurred, the
13
tumours are able to grow as a result of the new blood vessels that have formed. Once this
stage has been reached, the cancer cells can invade the blood vessel walls and spread to
other regions of the body through metastasis. Therefore, by providing the cells of the
tumour with access to nutrients and oxygen for growth, angiogenesis is a critical process
in tumour development and cancer progression.
Angiogenesis is controlled by both pro- and anti-angiogenic factors. The vascular
endothelial growth factor (VEGF) is the primary pro-angiogenic factor, for it binds to its
receptor (VEGFR) to initiate a signaling cascade that results in the activation of pathways
involved in permeability, survival, migration, and proliferation (Hoeben et al. 2004).
There are 7 members of the VEGF family (VEGFA-F and PIGF), and all share a
homologous domain located in exons 1-5 (Hoeben et al. 2004). The VEGFA gene is
located chromosomally at 6p12 and codes for a disulfide-linked homodimer that
functions as a glycosylated mitogen. This VEGFA gene contains 8 exons, and is
expressed in tissues of the adult lung, kidney, heart, and adrenal gland, and to a lesser
extent in the liver, spleen and gastric mucosa (Neufeld et al. 1999; Hoeben et al. 2004).
VEGFA gives rise to 7 different isoforms; 7 are splice variants, and 1 results from
proteolytic cleavage. VEGFB mRNA is predominantly expressed in the myocardium,
pancreas, and skeletal muscle and is composed of 7 exons that give rise to two splice
variants. For the scope of this thesis, these are the only two members of the VEGF
family that we will address. VEGF is the primary mediator of angiogenesis by playing an
14
important role in tumour vascularization and can be induced by hypoxic conditions, a
process that is mediated by HIF-1.
Tumour hypoxia occurs during tumour development and intermittently during
growth when the cells have outgrown their existing vasculature and cannot access
adequate oxygen from the circulation. During such periods, HIF-1 becomes active and is
able to activate the transcription of its target genes. VEGF contains hypoxia response
enhancer sequences in both its 5’ and 3’ domains and is coordinately upregulated under
hypoxic conditions (Maxwell et al. 1997; Dachs and Tozer 2000). Similar regions are
found in the erythropoietin (EPO) gene to increase the production of erythrocytes for
oxygen delivery (Semenza and Wang 1992; Semenza 1994; Hoeben et al. 2004). Thus
the hypoxia response, initiated by the HIF transcription factor, activates the transcription
of a variety of target genes involved in survival during hypoxia, including VEGF, the
primary moderator of angiogenesis.
1.4.3 Glucose uptake and glycolysis
One of the most striking indications of the hypoxia response is the induction of
genes involved in glucose metabolism in the cytoplasm. When oxygen is not available
and mitochondrial pathways are compromised, glycolytic metabolism must be induced to
provide the cell with energy (Guppy 2002). Most glucose transporters and glycolytic
enzymes are transcriptionally regulated by HIF-1. In terms of glucose uptake, HREs have
been identified in the promoters of the transporters GLUT1 and GLUT3 (Semenza et al.
1994; Ebert et al. 1995). Furthermore, Semenza et al. (1994) discovered that the
15
expression of aldolase A (ALDOa), phosphoglycerate kinase 1 (PGK1) and the muscle
isoform of pyruvate kinase (PKm) increased following exposure to hypoxia or chemical
hypoxia mimetics. Following this discovery, reporter constructs containing the promoter
regions of ALDOa, enolase 1 (ENO1), phosphofructokinase L (PFKl) and PGK1 were
activated in hypoxic cells, suggesting the presence of HREs (Firth et al. 1994; Semenza
et al. 1994; Firth et al. 1995). Further evidence by Semenza et al. (1996) confirmed that
activation of these glycolytic genes was HIF-1 mediated, as they were transcriptionally
upregulated in non-hypoxic cells that were overexpressing HIF-1α. The ability of the
transcription factor HIF-1 to regulate glucose uptake and glycolysis highlights the
importance of glycolysis in the cellular response to hypoxia.
1.4.4 Apoptotic resistance
The relationship between hypoxia and apoptosis is not well defined. While
evidence exists for hypoxia-induced apoptosis, in some instances hypoxia can confer
apoptotic resistance. The balance between these outcomes is controlled by HIF-1α and
other regulatory networks that respond to the cell’s microenvironment.
During periods of severe or prolonged hypoxia, necrosis or apoptosis generally
occur to prevent the accumulation of mutations that accompany environmental stress
(Greijer and van der Wall 2004). While there are both extrinsic (death receptor activated)
and intrinsic (mitochondrial) pathways of programmed cell death, environmental stress
and growth factor withdrawal trigger cell death using the latter of these two pathways.
During intrinsic apoptosis, DNA damage activates the transcription factor p53, which
16
stimulates cytochrome c release from the mitochondria by activating the apoptotic
proteins Bak and Bax (Wei et al. 2001). Following cytochrome c release and subsequent
binding with the apoptotic protease activating factor (Apaf-1), caspase 9 is activated (Wei
et al. 2001). Following this event, effector caspases and caspase substrates are cleaved
and hallmarks of apoptosis are apparent, including DNA fragmentation, chromatin
condensation, membrane blebbing, and formation of apoptotic bodies.
A reduction in the proton pumping by the electron transport chain during hypoxic
stress reduces ATP production and depolarizes the mitochondrial inner membrane. This
reduction in mitochondrial ATP is associated with the activation of the proapoptotic
proteins Bak and Bax, which stimulate cytochome c release from the mitochondria
thereby initiating the apoptotic cascade (Saikumar et al. 1998). Similarly, hypoxia may
induce apoptosis in a ROS-mediated process, which involves cleavage of caspase 9
independent of cytochrome c release (Kim and Park 2003). Kim and Park (2003)
proposed that this pre-activation of caspase 9 secondarily facilitates cytochrome c release
from the mitochondria by altering the mitochondrial membrane integrity. Furthermore,
HIF may play a role in hypoxia-induced apoptosis by stabilizing the tumour suppressor
gene p53, or by increasing the expression of the bcl-2 binding proteins, thereby inhibiting
the apoptotic effects of the protein. Therefore, periods of severe hypoxia or anoxia can
activate the intrinsic apoptotic pathway through multiple mechanisms.
While it seems evident that a cell would undergo apoptosis during hypoxia due to
either hypoxia-induced DNA damage or energy deprivation, what is more interesting is
17
the ability of some cancer cells to adapt to survive under such conditions. That is, when
hypoxia is acute or mild, cells that are able to adapt metabolically to their new
environment are able to evade the process of apoptosis. Perhaps cells that already express
HIF-1α at high levels under normoxia, due to hypoxia-independent activation of the HIF1 pathway, have a survival advantage when exposed to hypoxia in their
microenvironment. Hypoxia-independent HIF-1α protein accumulation can result from
genetic alterations such as mutations in proto-oncogenes or tumour suppressor genes. For
example, loss of VHL function can lead to HIF-1α stabilization under normoxia by
allowing the protein to escape proteasomal degradation. Pancreatic cancer cells have been
shown to overexpress HIF-1α protein under normoxic conditions, likely through
activation of the PI3K signaling pathway for cellular survival (Akakura et al. 2001).
Similar findings were shown in the prostate cancer cell lines, PC3, which express HIF-1α
constitutively under normoxic conditions (Zhong et al. 1998). Furthermore, PC3 cells are
androgen-independent, and thus represent a model of advanced stage prostate cancer.
Expression of HIF-1α can be correlated to tumor grade whereby more aggressive cancers
have greater responses to hypoxia (Hochachka et al. 2002). When cells that over-express
HIF-1α under normoxia are deprived of oxygen and glucose, they show more resistance
to apoptosis than do cancer cells that do not over-express HIF-1α under the same
conditions (Akakura et al. 2001). It is evident that HIF-1α determines whether a cell will
succumb to apoptosis or evade it in response to environmental hypoxia; however the
exact mechanisms underlying this ability have yet to be resolved. Perhaps cells that over18
express HIF-1α independently of hypoxia are able to survive and adapt when they are
exposed to hypoxia because the genes required for survival under hypoxic conditions are
already being transcribed.
1.5 Fatty acid synthesis and cancer
In addition to glycolysis, the TCA cycle, and the electron transport chain, fatty
acid synthesis is a crucial process to cancer metabolism. In lieu of the TCA cycle, cells
can shunt glucose to the pentose phosphate pathway (PPP), which plays a role in the
anabolism of fatty acids by producing reducing equivalents, and nucleic acids through the
production of their precursor, ribose 5-phosphate. De novo fatty acid synthesis is a
component of cellular anabolism and can be outlined as in Equation 1. 2:
Eq 1.2: citrate 
ACL
acetyl-CoA

malonyl-CoA
ACC

fatty acids (i.e. palmitate)
FAS
The overall reaction for fatty acid synthesis from acetyl CoA is shown in Equation 1.3:
Eq 1.3: acetyl-CoA + 7 malonyl-CoA +14 NADPH+ +14H+ 
palmitate + 7CO2 + 8 CoA +14 NAD+ + 6 H20
Citrate from the TCA cycle can be exported into the cytosol where it is converted
into acetyl-CoA by the enzyme ATP citrate lyase (ACL). Acetyl-CoA carboxylase (ACC)
then converts the acetyl-CoA into malonyl-CoA. The formation of malonyl-CoA is the
rate-limiting step of this process and can be controlled both allosterically and through
19
covalent modification. Its allosteric regulators include citrate, isocitrate, and αketoglutarate. Malonyl-CoA is then converted into palmitate by the enzyme fatty acid
synthase (FAS). This 16-carbon palmitate can be modified by chain elongation or
desaturation to produce other fatty acids via pathways in the mitochondria and
endoplasmic reticulum. However, free fatty acids do not accumulate in the cell under
normal circumstances and are rather converted into triacylglycerols for storage, or
phospholipids for cell membrane biosynthesis.
While aerobic glycolysis has been studied for decades, a more recently discovered
metabolic alteration in cancer cells is the up-regulation of the pathway of de novo fatty
acids, both at the mRNA and protein level. The link between fatty acid synthesis and
cancer was first discovered when an oncogene-marker, OA-519, was isolated from the
tumours of breast cancer patients and was later identified as the 270-kDa polypeptide,
fatty acid synthase (FAS) (Kuhajda et al. 1994). FAS is the final enzyme in the
endogenous fatty acid synthesis pathway that is involved in metabolizing glucose to fatty
acids. Since it was identified in breast cancer, overexpression of FAS protein has been
documented in colorectal, ovarian, endometrial, and prostate cancers (Rashid et al. 1997;
Hardman et al. 1995; Pizer et al. 1998; Shurbaji et al. 1992). The original proposal was
that this pathway was upregulated to support membrane biosynthesis in the cancer cells
as they undergo rapid cell division; however, there is a lack of evidence to support this.
Membrane biosynthesis seems an unlikely reason for upregulating fatty acid synthesis
because, while all cancer cells rapidly grow and divide, only certain types have an
20
upregulation of this pathway. Evidence indicates that FAS is regulated both
transcriptionally and posttranscriptionally (Rossi et al. 2003). FAS inhibitors have been
associated with apoptosis in cancer cells therefore suggesting that the pathway of fatty
acid synthesis is a potential target for cancer therapy (Kuhajda 2000). Furthermore, there
is a correlation between the overexpression of FAS and therapeutic resistance in cancer,
but this may merely be attributed to one of the other metabolic abnormalities that are
common in cancer (Swinnen et al. 2000).
1.5.1 Sterol regulatory element-binding protein (SREBP) transcription
factors
Sterol regulatory element-binding protein transcription factors (SREBPs) are the
master controllers of lipid metabolism. Among these basic helix-loop helix (bHLH)
leucine zipper transcription factors there are three isoforms that are produced from two
different genes. SREBP-1a and SREBP-1c are produced from the same gene (SREBF-1)
on chromosome 17 through alternative splicing in the first exon, and therefore only differ
in their NH2- terminal transactivation domains (Hua et al., 1995). SREBP-2 is produced
by a different gene (SREBF-2) on chromosome 22 but shares 47% homology with the
SREBP-1 protein (Miserez et al., 1997; Eberlé et al., 2004). SREBP precursor proteins
contain three structural domains: a) an NH2-domain for DNA binding and dimerization,
b) hydrophobic transmembrane spanning elements, and c) a COOH-terminal domain for
regulation (Eberlé et al., 2004; Ettinger et al., 2004). Both the NH2-domain and the
COOH-terminal components extend into the cytosol, and the transmembrane portion of
21
the protein anchors it in the ER membrane (Brown and Goldstein 1997). An overview of
the SREBP axis is depicted in Figure 1.
Figure 1. Possible ways that low
oxygen may influence the
SREBP axis via HIF-1α
The influence of low oxygen on
the SREBP axis may be mediate
by the transcription factor HIF-1α.
Directly, HIF-1α may bind to
HREs in the promoters of
lipogenic genes. Alternatively,
HIF-1 may be indirectly involved
in de novo fatty acid synthesis by
exerting its effects on the
regulation of SREBP transcription
factors. (Figure adapted from
Eberlé et al., 2004).
SREBPs are synthesized as precursor proteins that are bound by the endoplasmic
reticulum and stabilized by sterol levels within the cell. Here they associate with the
SREBP cleavage activating protein (SCAP), and this complex is retained in the ER
membrane by Insig proteins. When sterol levels are low or insulin is present, SCAP is
released from Insig, and is able to transport SREBP to the Golgi apparatus. Here it
undergoes two cleavage reactions, the first by a site-1 protease to cleave the ER luminal
component, and the second by a site-2 protease, which releases the NH2-domain. The
NH2-domain becomes the active version of the protein containing the bHLH-LZ domain.
This component is then be translocated to the nucleus (nSREBP) where it dimerizes to
22
activate the transcription of its target genes, which contain sterol regulatory elements
(SRE) or E-box sequences in their promoter regions.
1.5.2 Androgens and lipogenesis
Androgens have been known to play an important role in the normal and
cancerous development of the prostate for many years. The survival of prostate cells
relies on the androgen receptor, which, upon nuclear localization in the presence of
androgens, activates the transcription of its target genes involved in cell growth (Debes
and Tindall 2004). As a result, androgen deprivation therapy has been a common method
of treatment for metastatic prostate cancer; however, it does not prevent the progression
towards androgen-independence (Shaw et al. 2007). Interestingly, studies have
demonstrated that even androgen-independent prostate cancers maintain androgenreceptor (AR) activity (Chen 2004). AR gene amplification (Visakorpi et al. 1995), AR
mutations (Taplin et al. 1995), or the activation of the AR by non-androgen ligands
(Culig et al. 1994; Lee et al. 2003) may explain the active androgen signaling that occurs
in refractory prostate cancer. These alterations are not commonly detected in patients
with refractory disease (Chen et al. 2004). Alternatively, AR activity may persist
following androgen ablation therapy due to the actual presence of androgens. Mohler et
al. (2004) detected levels of dihydrotestosterone sufficient to activate the AR in tissues of
recurrent prostate cancer. Androgen-independence is associated with a decrease in
median survival and poor medical prognosis (Petrylak et al. 2004; Shamash et al. 2005).
23
The molecular mechanisms underlying the survival advantage these cancer cells have
likely involves androgen signaling.
The interplay between hormonal regulation and metabolism has yet to be
uncovered. Not only are they involved in cellular growth and survival, but androgens
have also been shown to cause an accumulation of neutral lipids in the androgen-sensitive
LNCaP cell line (Swinnen et al. 1996a). Swinnen et al. (1997) demonstrated that the
synthetic androgen, R1881, was able to induce an upregulation of FAS mRNA and
enzyme levels as well as an accumulation of lipids in LNCaP cells. The induction of fatty
acid synthesis LNCaP cells in response to androgens is mediated by SREBP transcription
factors (Swinnen et al. 1997). This was supported by Heemers et al. (2001), who
discovered that androgens increase the abundance of nSREBPs in two androgen
independent lines, LNCaP and MDA-PCa-2a. It has been proposed that this androgenmediated increase in fatty acid synthesis may occur in response to hypoxia (Park et al.
2006).
1.5.3 Hypoxia and fatty acid synthesis
While androgens seem to induce the upregulation of the pathway of fatty acid
synthesis, it has yet to be determined how this fits in with a cell’s overall metabolism.
Hochachka et al. (2002) suggested that the fatty acid synthesis pathway might be
upregulated in prostate cancer to maintain cellular redox balance when oxygen is lacking.
During glycolysis and the TCA cycle, reducing equivalents are produced in the form of
NADH and FADH2, which are shuttled to the inner membrane of the mitochondria. In the
24
mitochondria, the energy from their oxidation is used to pump protons into the
intermembrane space, and the electrons are eventually transferred to oxygen, the final
electron acceptor. There are two possible ways in which a cell can maintain redox
balance under hypoxic conditions when reducing equivalents cannot be oxidized via
mitochondrial pathways: 1) using LDH to catalyze the formation lactate, or 2)
synthesizing fatty acids. Thus the pathway of fatty acid synthesis may be upregulated as a
means to maintain redox balance within the cell during periods of hypoxia, when the
respiratory chain is unable to do so.
The process of lactate formation from pyruvate occurs in the cytoplasm and
represents one mechanism by which a cell may maintain redox balance in the absence of
oxygen. For every molecule of glucose entering glycolysis, 2 molecules of pyruvate are
generated. Rather than forming acetyl-CoA, a process catalyzed by PDH, under anaerobic
conditions, the two molecules of pyruvate can be reduced to lactate by LDH. This process
oxidizes the 2 NADHs that are generated in glycolysis during the conversion of
glyceraldehyde phosphate into 1,3-diphosphoglycerate. Therefore by oxidizing the
NADHs produced in glycolysis, the process of lactate production serves as a way which
the cell can maintain redox balance when its mitochondrial pathways are compromised.
Alternatively, the pathway of fatty acid synthesis may serve to maintain redox
balance in the hypoxic cell. In addition to the 2 NADHs formed in glycolysis, many
reducing equivalents are produced during the TCA as well. The formation of acetyl-CoA
from pyruvate, α-ketoglutarate from isocitrate, succinate from α-ketoglutarate, and
25
oxaloacetate from malate all produce NADH. Additionally, FADH2 is produced during
the oxidation of succinate into fumarate. Thus, per molecule of glucose that enters the
glycolytic pathway and the TCA cycle, 10 NADH and 2 FADH2 are produced. In the
absence of oxygen, this results in a major redox imbalance, for, without the final electron
acceptor in the electron transport chain, the electrons from these carriers transport will
have nowhere to go and the cell will maintain in a reduced state. For every seven
molecules of glucose 6-phosphate converted into ribulose 5-phosphate via the PPP, a
sufficient amount of NADPH is formed to produce 1 molecule of palmitate. Recall, 14
NADPH are required for the formation of palmitate via the pathway de novo fatty acid
synthesis. It therefore seems unlikely that the LDH is responsible for maintaining redox
balance in the cell alone, for it only oxidizes 2 NADHs per molecule of lactate produced.
The pathway of fatty acid synthesis may play a larger role, for the reducing power of 14
NADHs is required for the formation of one palmitate. This pathway requires much more
reducing power than the process of lactate production, and may serve to maintain the cell
in a more favorable oxidized state when oxygen is limiting (Hochachka et al. 2002).
The precise molecular mechanisms that drive the upregulation of the de novo fatty
acid synthesis pathway in cancer cells are not known. The relationship between hypoxia
and androgen signaling pathways may be of importance. The ability of the AR to respond
to such low levels of androgens following androgen ablation therapy may in part be
controlled by the HIF-1 mediated hypoxia signaling pathway. Park et al. (2006) showed
that hypoxia increased AR-ARE binding, expression of AR controlled genes, and
26
translocation of AR to the nucleus. This suggests that, in order to remain redox balance in
the cell, that hypoxia signaling may indirectly lead to the activation of the fatty acid
synthesis pathway through an AR signaling mechanism. However, whether there is a
direct relationship between hypoxia and fatty acid synthesis that is androgen-independent
has yet to be determined.
1.6 Objectives
My thesis focuses on elucidating the reasons for and molecular mechanisms that
enable prostate cancer cells to survive and progress under hypoxic conditions. This
involves a precise characterization of cellular metabolism and the influence of hypoxia.
While cultured cells are commonly used as models to study cancer, metabolic differences
between cell lines and how they behave in vitro compared to in vivo must also be
considered. The objectives of this thesis are outlined below.
Objective 1: Determine the origin of the metabolic phenotype of prostate
cancer cells. Cancer cells have an unusual metabolism, favoring glycolysis over
oxidative phosphorylation for ATP production, despite reduced ATP yields and
independent of oxygen levels. The reasons for this pattern remain unclear, but may be a
result of mitochondrial dysfunction. Researchers often attribute the phenotype of aerobic
glycolysis to all cancer cells without proper experimental evaluation. Thus, my first
objective is to establish the metabolic phenotype of 3 human prostate cancer cell lines:
LNCaP, DU145, and PC3. I hope to determine which energy pathway is predominate for
meeting cellular ATP demands, and looking further into the foundations of their
27
metabolism. I hypothesize that each of these cell lines will exhibit a similar metabolic
phenotype in terms of the enzymes of energy metabolism and relative reliance on
glycolysis versus oxidative phosphorylation for energy.
Objective 2: Determine the influence of passage and selection on the
metabolic phenotype of prostate cancer cells. It is generally accepted that cells cultured
for many generations in vitro can evolve to become different from the original cells.
Although LNCaP cells are perhaps the most common cell line used in studies of prostate
cancer, however very little is known about the metabolic phenotype of the LNCaP line
and its numerous derived lines. As well, these cells are commonly used to study tumour
properties in vivo through xenograft approaches. Thus, my second objective is to explore
the influence of growth context (passage number, in vitro versus in vivo growth) on the
metabolic phenotype of LNCaP-derived lines. I hypothesize that the cell line will show
changes in metabolic phenotype in relation to both passage number and growth context.
Objective 3: Determine the relationship between hypoxia and fatty acid
synthesis. Many cancers, including prostate cancer, have unusually high rates of fatty
acid synthesis. This ability is associated with and contributes to tumour progression. It
may serve to supply phospholipids to membranes or maintain redox balance in the cell
during hypoxia. Whether this pathway is also responsive to hypoxia is unclear. Thus, my
third objective is to explore the influence of oxygen levels on the expression of genes in
the SREBP axis and fatty acid synthesis pathway, and to determine whether or not
hypoxia influences the pathway at a transcriptional level. Furthermore, I set out to
28
quantify TG production in prostate cancer cells following different low oxygen
treatments to determine if the previously described upregulation in FAS mRNA and
enzyme levels follows through to the accumulation of lipids in these cells. I hypothesize
that the pathways of fatty acid synthesis – from the SREBP axis to TG biosynthesis – is
neither hypoxia sensitive nor involved in hypoxic redox balance.
29
Chapter 2: Hypoxia and the metabolic phenotype of prostate cancer cells
2.1 Introduction
Otto Warburg was the first to reveal that, even when normoxic, many cancers rely
on glycolysis rather than oxidative phosphorylation as a primary means of ATP synthesis
(Warburg 1930). This preferential use of glycolytic metabolism may arise from growth
and selection in an hypoxic microenvironment, and may lead to a more aggressive, less
responsive cancer (Vaupel et al. 2001; Ghafar et al. 2003; Acs et al. 2004). The
underlying explanations for aerobic glycolysis and its precise relationship with hypoxia,
however, are not completely understood. The ability to survive when oxygen is limiting
is controlled in part by hypoxia-inducible factor (HIF), the transcription factor of an
oxygen-sensitive pathway that regulates genes involved in anaerobic metabolism,
angiogenesis, growth factor signaling and metastasis (Semenza et al. 1996; Coffey et al.
2005; Carmeliet et al. 1998). If, as suggested, the glycolytic phenotype is due to growth
in hypoxia, it is unclear why this phenotype persists following vascularization (reviewed
by Costello and Franklin 2005). It is also possible that the glycolytic phenotype is a
compensatory response to a reduction in mitochondrial content or function (Warburg
1930).
Glycolysis serves many roles in anabolic metabolism (e.g., ribose and NADPH
synthesis), but the main role of glycolysis in cancer cells is thought to be ATP
production. The NADH produced in glycolysis is normally oxidized by mitochondria via
30
redox shuttles, but under hypoxic conditions, the NADH must be oxidized by other
routes. When mitochondrial metabolism is insufficient, most mammalian cells produce
lactate to oxidize NADH. However it has been suggested that cancer cells may also rely
on fatty acid synthesis to balance redox (Hochachka et al. 2002). Transhydrogenation
reactions convert cytoplasmic NADH to NADPH, which is used by fatty acid synthase
(FAS). Though prostate and many other cancers accumulate high intracellular levels of
lipid through de novo fatty acid synthesis (Medes et al. 1953; Swinnen et al. 1996b;
Moreau et al. 2006; Ookhtens et al. 1984), this capacity is more often linked to the need
to support membrane biosynthesis in the rapidly dividing cells (reviewed by Hochachka
et al. 2002; Swinnen et al. 2003).
The role of fatty acid synthesis in cancer cells remains unestablished, though it is
clearly important. Many cancer cells show an up-regulation of FAS mRNA and protein,
as well as a greater expression of other members of the pathway of de novo fatty acid
synthesis (Shurbaji et al. 1996; Swinnen et al. 2002; Welsh et al. 2001; Rossi et al.
2003). When the expression of acetyl CoA carboxylase (ACC) or FAS is blocked,
prostate cancer cells show reduced proliferation rates and ultimately apoptotic death,
suggesting that these enzymes may be potential targets for anti-cancer therapy (Pizer et
al. 1996; Brusselmans et al. 2005). It is not yet clear if these enzymes of lipogenesis, like
those of glycolysis, are responsive to hypoxia. The profile of enzymes that control lipid
homeostasis (e.g., ACC1, FAS) is regulated in large part by sterol regulatory element
binding protein (SREBP) transcription factors (Eberlé et al. 2004).
31
In the present study, we assess how the enzymes of intermediary metabolism
(mitochondrial, glycolytic, and lipogenic) and their genetic regulators differ among
prostate cancer lines (LNCaP, DU145, and PC3). We also reconcile the relationship
between the metabolic profile and line-specific responses to hypoxia. Focusing on
LNCaP and derived lines, we assess how passage/clonal selection affects the metabolic
phenotype both in vitro (in culture) and in vivo (in xenografts). Collectively these studies
determine the relationship between aerobic glycolysis, tumour hypoxia, and lipid
metabolism in prostate cancer cells with the goal of gaining a better understanding of the
origins of the aerobic glycolytic phenotype of prostate and other cancer cells.
2.2 Methods
2.2.1 Cell culture and treatments
LNCaP, PC3, and DU145 prostate cancer cell lines were obt
ained from the American Type Culture Collection (ATCC, Manassas, VA).
LNCaP was derived in 1977 from a needle aspirate biopsy of a supraclavicular lymph
node from a 50-year-old white male with stage D prostatic cancer (Horoszewicz et al.
1980). LNCaP cells express a mutated form of the androgen receptor, which leads to
some alterations in androgenic responses (Veldscholte et al. 1992). However, these cells
produce the human prostatic secretory markers prostate specific antigen (PSA) and
prostatic acid phosphatase (PAP) both in vitro and when xenografted into nude mice
(Chung et al. 1992). LNCaP cells are responsive to androgens in terms of growth
32
however they exhibit aberrant responses to antiandrogens (Wilding et al. 1989). The PC3
line was isolated from a bone metastasis of a 62-year-old Caucasian diagnosed as having
undifferentiated grade IV adenocarcinoma of the prostate. PC3 will grow in soft agar, in
suspension culture and will form tumours in nude mice. It does not respond to androgens
or various growth factors (Kaign et al. 1979; Kaign et al. 1980). DU145 cells were
derived in 1975 from a brain metastasis of prostatic cancer in a 69-year-old white male.
The original metastasis was identified as moderately differentiated with foci of poorly
differentiated cells. DU145 cells grow very rapidly from low densities. They are neither
hormone dependant nor hormone sensitive in terms of growth (Kaign et al. 1979; Mickey
et al. 1980).
LNCaP-luc were a gift from Dr. Mark Day (University of Michigan). They were
derived from LNCaP cells that had been transfected with luciferase constructs and
clonally selected. High passage LNCaP cells (HP-LNCaP) were derived by continuous in
vitro culture of LNCaP cells through approximately 50 passages. To mitigate potential
effects of differences in culture media, we grew all cells in Dulbucco’s Modified Eagle
Medium (DMEM: high glucose with glutamine, Gibco, Burlington, ON) supplemented
with 10% fetal bovine serum (FBS, Gibco, Burlington, ON) and 1% penicillinstreptomycin (Gibco, Burlington, ON), with 5% CO2 and 95% humidity at 37˚C. Acidic
medium was obtained using the standard DMEM mixture replacing 24 mM NaHCO3
with a mixture of 0.94 mM NaHCO3 and 22 mM NaCl.
33
Hypoxia treatments (2% O2) were performed either in a modular incubator
chamber (Billups-Rothenberg Inc., San Diego, CA) flushed with 2% O2, 5% CO2 and
balance nitrogen, or in an incubator with 5% CO2 and 2% O2 at 37˚C. Anoxia treatments
(0% O2) were performed in a chamber flushed with 5% CO2 and balance nitrogen. For
treatments with azide, a chemical anoxia mimetic, a final concentration of 5 mM was
added to the plates. For sodium dichloroacetate (DCA) studies, cells were treated with 0.5
mM of the drug. For 2,4-dinitrophenol (DNP) studies, cells were exposed to a final
concentration of 0.1 µM of DNP prior to measurements.
2.2.2 Oxygen consumption and lactate production measurements
Respiration was measured in cell suspensions using Clarke-type electrodes fitted
into water-jacketed vessels (Gilson, Middleton, WI). Cells were trypsinized, centrifuged
(3 min at 200 g) and resuspended in media that was air-saturated at 37oC.
To assess the impact of oxygen concentration on lactate production, cells were
exposed to control (21% O2, 5% CO2, balance N2), hypoxia (2% O2, 5% CO2, balance
N2), anoxia (5% CO2, balance N2), or azide (5 mM) for 12 hours. Cells were collected
from the supernatant and the plate, and aliquots were centrifuged (12,000 g for 1 min).
The supernatant was analyzed for lactate concentration using the hydrazine-sink method
(see Moyes et al. 1997). In brief, media and lactate standards were incubated in 96-well
plates in the presence of NAD+ (2 mM), lactate dehydrogenase (1 U/ml) and a glycinehydrazine buffer (pH 9.2). Lactate production rates were expressed relative to the cellular
protein collected from the medium and adherent cells.
34
2.2.3 Enzyme assays
Cells were trypisinized and then centrifuged (3 min at 200 g). Pellets were washed
with phosphate-buffered saline (PBS, 0.731 M NaCl, 0.037 mM KCl, 0.022 mM
Na2HPO4⋅7H2O, 0.087 mM K2HPO4, pH 7.4), flash frozen in liquid nitrogen and stored
at -80ºC prior to enzyme analyses. The cell pellets were resuspended in enzyme
extraction buffer (20 mM HEPES, pH 7.4, 0.1% Triton X-100, 1 mM EDTA) prior to
spectrophotometric assays. The activities of mitochondrial cytochrome oxidase (COX)
and citrate synthase (CS) were determined as previously described (Moscow et al. 1988).
For lactate dehydrogenase (LDH), pyruvate kinase (PK), and glyceraldehyde 3-phosphate
dehydrogenase (G3PDH) activity, the absorbance of NADH (340 nm) was measured over
3 minutes. The assay buffers were as follows: LDH (50 mM HEPES pH 7.4, 0.2 mM
NADH, and 2 mM pyruvate), PK (50 mM HEPES pH 7.4, 0.1 M KCl, 10 mM MgCl2,
0.2 mM NADH, 5 mM phosphophenolpyruvate, 5 mM ADP pH 6.8, 10 µM fructose-1,
6-bisphosphate, and 10 U/ml LDH), and G3PDH (40 mM bicine, 0.8 M sodium acetate,
0.8 mM EDTA, 25 mM sodium arsenate, 0.5 mM NAD+, and 1 mM glyceraldehyde 3phosphate). Hexokinase (HK) and phosphoglucoisomerase (PGI) assays were linked to a
glucose 6-phosphate dehydrogenase (G6PDH) catalyzed reaction. The buffers for these
assays were as follows: HK (50 mM HEPES pH 7.4, 5 mM MgCl2, 2 mM NAD+, 0.5
mM NADP+, 1 mM ATP, 5 mM dithiothreitol, 5 mM glucose, 10 U/ml glucose 6phopshate dehydrogenase), and phosphoglucoisomerase (50 mM Tris pH 8.1, 0.5 mM
NADP+, 5 mM dithiothreitol, 0.5 mM fructose 6-phosphate, and 10 U/ml glucose 635
phopshate dehydrogenase. All enzyme analyses were normalized to the concentration of
protein in the samples.
2.2.4 RNA extraction and real-time PCR
Cells were homogenized in buffer RLT using a 21-gauge needle and RNA was
extracted using the RNeasy extraction kit (Qiagen, Mississauga, ON). Complimentary
DNA was synthesized using AMV RT (Promega, Madison, WI), oligo (dT) primers, and
random hexamers as per manufacturer’s instructions. Real-time PCR was performed
using 50 ng of cDNA, 0.58 µM of the appropriate primers (Table 1), 12.5 µl of SYBR
green (Qiagen, Mississauga, ON) and 3.5 µl of water. An Applied Biosystems 7500 realtime PCR system (ABI, Foster City, CA) was used under the following conditions: a
95ºC denaturation phase for 15 minutes followed by 40 cycles consisting of 15 seconds at
95ºC, 30 seconds at the optimal annealing temperature and 36 seconds at 72ºC. All
primers were human-specific and were tested on mouse tissue to ensure no amplification
would occur due to residual mouse tissue from xenografts. Primers were designed to
flank intron-exon boundaries whenever possible to avoid amplification of genomic DNA.
All expression levels are relative to the expression of the housekeeping gene TATAbinding protein (TBP).
36
Table 1. Real-time PCR human specific primers (5’-3’)
Gene (Accession No.)
Forward primer
Reverse Primer
TBP (NM003194)
cagtgacccagcagcatcact
aggccaagccctgagcgtaa
HIF-1α1 (NM001530)
ctagccgaggaagaactatgaacat
ctgaggttggttactgttggtatca
VEGF65 (NM001025366)
ccctgatgagatcgagtacatctt
agcaaggcccacagggattt
VEGF212 (NM001025367)
ccctgatgagatcgagtacatctt
gcctcggcttgtcacattt
SREBP-1a (NM004176)
aggcggctttggagcag
agcatgtcttgtcacattt
SREBP-1c (NM001005291)
gccatggattgcacttt
caagagaggagctcaatg
SREBP-2 (NM004599)
cccctgacttccctgctgca
gcgcgagtgtggccggatc
SCAP (NM012235)
tgctcaccgtggggatgt
cactgctgatgacacaggaggt
FAS (NM004104)
caagcaggcacacacgatgg
ggtctcggctcagggcctcc
ACC1 (NM001093)
ctgacgaggactctgttgc
ggtggagtcccgacatgct
HKII (NM000189)
acggagctcaaccatgaccaa
ccatctggagtggacctca
LDHa (NM005566)
tggcctgtgccatcagtatct
gccgtgataatgaccagcttg
PKm (NM002654, NM182470,
caccgcaagctgtttgaa
tgccagactccgtcagaact
ctggctgctgtctacaaggct
cctcctcactctggcctcc
PGC-1α (NM013261)
caggtatgacagctacgaggaa
tgcctctgcctctcccttt
NRF-2α (NM002040)
ttccagcatcagtgcaatct
ctgaaatcctcggcgctct
COXIV2 (XM008055)
cctcctggagcagcctctc
tcagcaaagctctccttgaactt
COXI3 (NM536845)
ttcgccgaccgttgactattctct
aagattattacaaatgcatgggc
CS (NM004077, NM198324)
gcctgtacctcaccatccca
tttgccaacttccttctgca
p53
ctgctcagatagcgatggtctg
ttgtagtggatggtggtacagtca
SCO2
cttcactcactgccctgaca
cggtcagacccaacagctt
NM182471)
ALDOa (NM000034,
NM184041, NM184043)
1
Kim et al. (2005); 2 Stump et al. (2003); 3 Masayesva et al. (2006)
37
2.2.5 Intracellular ATP levels
After growing cells to 80% confluency, the medium was replaced and cells were
exposed to control, acidic (pH 6.2), hypoxic, or acidic and hypoxic conditions for 72
hours. Cellular ATP levels were determined in acid extracts of cells. After treatment, the
supernatant was collected and pooled with trypsinized cells. After removing 10% of the
cells for protein determination, the remaining cells were sedimented (5 min at 200 g) and
the pellets were acidified (0.5 ml of 3% perchloric acid). After centrifugation, 400 µl of
the supernatant was neutralized with 50 µl saturated Tris base, 50 µl 2M KCl, and 150 µl
2 M KOH with trace amounts of phenol red as a pH indicator. Extracts were analyzed for
ATP using a luciferin-luciferase assay mixture (Sigma) on an LMax luminometer
(Molecular Devices, Sunnyvale, CA). The intracellular ATP levels were then normalized
to the protein concentration in each sample.
2.2.6 Cardiolipin content (NAO staining)
Mitochondrial content was determined by staining the cells with nonyl acridine
orange (NAO) (AnaSpec, San Jose, CA), a dye that quantifies mitochondrial mass by
binding to cardiolipin in mitochondria, regardless of their energetic state (Dement et al.
2007). In brief, cells were grown in clear medium, and were treated with 100 nM NAO
for 30 min. Following treatment, cells were washed, trypsinized with clear trypsin,
centrifuged for 5 minutes at 500g, and resuspended in PBS.
38
2.2.7 Xenografts in SCID mice
For sub-renal capsule grafting of LNCaP cells, 1-5 x 105 cells were embedded in
50 µl rat tail collagen as described previously (Hayward et al. 1999). After an overnight
incubation at 37oC, the collagen gels were grafted under the kidney capsule of
testosterone-supplemented, castrated adult male severe combined immunodeficient
(SCID) mice (Harlan, Indianapolis, IN). LNCaP-luc tumours were generated by intrafemoral injections of twenty five thousand cells into adult male SCID mice. Tumours
formed from outgrowths from the injection sites in the femur, into the surrounding leg
muscle. After 28 days, all mice were sacrificed and tumours harvested and snap-frozen in
liquid nitrogen.
2.2.8 Triglyceride extraction and quantification
Lipids were extracted from prostate cancer cells using a modification of the Folch
method (Folch et al. 1957). Briefly, cell pellets were dissolved in chloroform: methanol
(2:1) and were centrifuged for 10 minutes at 3000 g. The resulting mixture was then
filtered, and 0.25% KCl was added to remove the water-soluble compounds. Following
separation of the two phases and removal of the aqueous phase by vacuum, samples were
evaporated under nitrogen gas at 60ºC and resuspended in isopropanol. The resulting
lipids were then quantified using the TRIGS kit (Randox, Muissisauga, ON). This kit
uses a colorimetric method to quantify lipids following their enzymatic hydrolysis with
lipases.
39
2.2.9 Statistical analyses
Analysis of variance (ANOVA) tests were used to determine statistical
significance. First, each treatment was tested for both normality using the Shapiro-Wilk
test and for equal variance using the O’Brien test. If the results from both of these tests
were not statistically significant (p<0.05) then a one-way ANOVA was performed
followed by a Tukey HSD test. If the data were not normal or equally distributed, then
the non-parametric Wilcoxon test was used. For dichloroacetate studies, a dependent
Student’s t-test was used to determine if the effects of the drug were significant following
treatment. All tests were performed using JMP statistical software. For our experiments,
any significantly different values have a p value of less than 0.05.
2.3 Results
2.3.1 The metabolic phenotype differs between cell lines
To determine metabolic differences between prostate cancer cell lines, we
evaluated rates of oxygen consumption and lactate production, and levels of ATP. We
also explored how hypoxia and low pH influenced their respective metabolic patterns.
The metabolic phenotype of LNCaP cells differed from that of DU145 and PC3
cells; LNCaP cells had a significantly greater oxygen consumption rate (Figure 2A), as
well as a significantly lower rate of lactate production (Figure 2B). The relative
importance of OXPHOS and glycolysis can be estimated from these data, assuming 1
nmol O2 consumed equates to 5 nmol ATP produced, and 1 nmol lactate produced
40
equates to 1 nmol ATP produced. With these assumptions, OXPHOS contributed 88% of
ATP demands in LNCaP cells, but only 56% in PC3 and 47% in DU145 cells.
Figure 2. Metabolism of LNCaP,
PC3, and DU145 cells
Basal metabolic measurements for
prostate cancer cells in terms of A)
O2 consumption and B) lactate
production. Bars are means (+SEM)
with different letters denoting
significantly different values
between cell lines according to
ANOVA and Tukey tests (n=5).
The three cell lines also differed in their response to oxygen supply and OXPHOS
inhibition (Figure 3). LNCaP cells were most affected, increasing their lactate production
rate nearly 3-fold under hypoxia and about 4-fold with either anoxia or azide treatment.
The PC3 cells also showed an increase in lactate production with these treatments, though
the maximal stimulation was only about 2-fold. DU145 cells significantly increased their
production of lactate following treatment with anoxia and azide only. The greater
sensitivity of LNCaP cells to OXPHOS limitations was also reflected in ATP levels
(Figure 4). In addition to low oxygen treatment, I included an acidic treatment. In cells
exposed to hypoxia, here is both reduced oxygen availability and an acidosis that arises
from the Pasteur effect (elevated rates of lactate production). LNCaP cells had 4-fold
41
higher ATP levels than DU145 and PC3 cells under normoxic conditions, and responded
much more dramatically to low oxygen treatments (Figure 4). When exposed to hypoxia,
ATP levels declined by 50% in LNCaP cells, but were not affected in either PC3 or
DU145 cells.
Figure 3. Lactate production.
Cells were exposed to normoxia,
hypoxia, anoxia, or azide for 12
hours. Bars are means (+SEM)
with different letters denoting
significantly different values
within each cell line according
to ANOVA and Tukey tests
(n=5).
42
Figure 4. ATP levels following
exposure to low pH and hypoxia
Cells were exposed to low pH
(6.2), hypoxia (2% O2), or the
combination of low pH and
hypoxia for 72 hours. Bars are
means (+SEM) with different
letters denoting significantly
different values within each cell
line according to ANOVA and
Tukey tests (n=5).
The ability of these cells to respond to dichloroacetate (DCA) was also assessed.
This inhibitor of pyruvate dehydrogenase kinase stimulates pyruvate dehydrogenase
(PDH), consequently glucose oxidation (Bonnet et al. 2007). DCA increased the oxygen
consumption rate in LNCaP cells, but had no effect on DU145 or PC3 respiration rates
(Figure 5A). Moreover, DCA did not significantly influence the lactate production rates
in any of the cell lines evaluated (Figure 5B). Collectively, these studies suggest that
LNCaP cells have normal mitochondrial metabolism but both the PC3 and DU145 cells
suffer some degree of mitochondrial impairment that accompanies glycolytic stimulation.
43
Figure 5. Metabolism of LNCaP, DU145, and PC3 cells following treatment with
dichloroacetate (DCA)
A) O2 consumption, and B) lactate production. Bars represent means (+SEM) and are
scaled to the control for each cell lines. * Indicates significant difference according to a
Student’s t-test (n=6).
2.3.2 DU145 and PC3 cells have reduced mitochondrial enzyme activity but
not mitochondrial gene expression
Next, the relative activities of selected marker enzymes of glycolysis and
oxidative metabolism were assessed (Figure 6). Although there were no consistent
enzymatic differences in the glycolytic pathways in the three cell lines, DU145 and PC3
cells possessed less than half the activities of the mitochondrial enzymes COX and CS.
44
Figure 6. Glycolytic and
mitochondrial enzyme activities
in DU145 and PC3 cells
Bars are means (+SEM) with
different letters denoting
significantly different values for
each enzyme assayed according to
ANOVA and Tukey tests (n=6). All
enzyme levels are expressed
relative to LNCaP means: PK 4.3 ±
0.12, LDH 2.9 ± 0.2, PGI 2.7 ± 0.1,
G3PDH 1.2 ± 0.03, HK 0.1 ±
0.007, COX 0.013 ± 0.001, CS
0.386 ± 0.015 enzyme units
(nmoles/min/mg protein).
Transcript abundance was then explored to determine if there were differences in
the mRNA levels of metabolic genes and their regulators (Figure 7). There were
pronounced differences in the expression of genes linked to the hypoxic response (Figure
7A). PC3 cells showed elevated mRNA levels of a broad suite of hypoxia-related genes
including HIF-1α (30-fold), two VEGF isoforms (30-fold), and each glycolytic gene we
studied (5- to 10-fold). DU145 had intermediate transcript abundance, showing similar
elevations in HIF-1α and VEGF transcripts, but without the elevations in the expression
of glycolytic enzymes.
There were also differences in the transcript levels of genes linked to
mitochondrial metabolism (Figure 7B). Despite lower mitochondrial flux (Figure 2) and
lower mitochondrial enzyme activities (Figure 6), both DU145 and PC3 cells showed an
apparent stimulation of multiple mitochondrial enzymes (Figure 7B). DU145 cells
45
possessed more abundant transcripts for COXIV, COXI, CS, and the COX assembly
factor SCO2, whereas PC3 cells had high NRF-2 and SCO2 transcript levels.
Although there were remarkable line-specific differences in both glycolytic and
mitochondrial gene expression, there were only modest differences in the expression of
genes linked to fatty acid metabolism (Figure 7C). The two glycolytic cell lines had a
significantly greater expression of the SREBP-1a, SREBP-2, and SCAP but lower
expression of SREBP-1c, the isoform linked to lipogenesis. There were no significant
differences cell lines in FAS expression, but the expression of ACC1 was significantly
greater in the DU145 and PC3 cells.
46
Figure 7. mRNA levels in LNCaP,
DU145, and PC3 cells
A) Hypoxia related mRNAs, and B)
mRNAs for mitochondrial genes and
transcriptional regulators. All transcript
levels are displayed in relation to LNCaP
mRNA levels. Bars are means (+SEM) with
different letters denoting significantly
different values between cell lines according
to non-parametric Wilcoxon tests (n=6). See
Supplementary Table 1 for raw cycle
thresholds relative to TBP.
To confirm our hypothesis that the glycolytic cell lines DU145 and PC3 suffered
mitochondrial impairment, we tested the ability of DNP, an oxidative uncoupler, to
stimulate mitochondrial respiration. We discovered that DNP was able to increase oxygen
47
consumption in the LNCaP cells by 112%. DNP increased the rate of cellular respiration
less drastically in the PC3 cells by 42% (Figure 8). No significant increase in oxygen
consumption resulted in the DU145 cells in the presence of DNP (Figure 8).
Figure 8. Oxygen consumption of
LNCaP, DU145 and PC3 cells
following exposure to
dinitrophenol (DNP)
Bars represent mean oxygen
consumption (+SEM) in the
presence of 0.1 uM DNP. *Indicates
significance according to a Student’s
t-test (n=5).
We used NAO, a fluorescent ligand of the mitochondrial phospholipid cardiolipin to
assess if the differences in mitochondrial enzyme activities were simply differences in
mitochondrial content (Figure 9). None of the cell lines showed differences in cardiolipin
content, suggesting LNCaP cells had the same number of mitochondria, but higher
mitochondrial enzyme specific activities. Given the lower respiration rates and enzyme
activities (yet similar mitochondrial inner membrane contents), this suggests that the
depolarization is these lines was due not to uncoupling but rather less proton pumping.
48
Figure 9. NAO fluorescence in LNCaP,
DU145, PC3, and LNCaP-derived cell
lines measured using FACS
Median fluorescence levels are displayed in
relation to the maximum value for each
run. Bars are means (+SEM) (n=6). Results
represent the median value taken from 5
separate experiments
2.3.3 LNCaP cells transition to a glycolytic phenotype with increasing
passage and clonal selection
We studied two lines of LNCaP cells that had been subjected to repeated passage
number (HP-LNCaP) or additional clonal selection (LNCaP-luc). We also used these
cells to assess the plasticity of the metabolic phenotype. In both lines, prolonged passage
under normoxia reduced the rates of oxygen consumption (Figure 10A) and increased the
rate of lactate production (Figure 10B). The high passage lines achieved a metabolic
phenotype similar in most respects to the PC3 and DU145 cells. Both lines showed a
significant “hypoxic-response” in terms of elevations in expression of HIF-1α and VEGF,
as well as selected glycolytic genes (Figure 11A). HIF-1α and PKm mRNA levels were
dramatically higher in LNCaP-luc cells but less so in HP-LNCaP cells (Figure 11A).
There was no uniform pattern seen in mitochondrial gene expression with
passage. HP-LNCaP cells had a significantly higher expression of PGC-1α, NRF-2 and
49
COX1; LNCaP-luc cells had elevated expression of NRF-2, COXIV and COXI (Figure
11B). There were also differences seen in the SREBP axis (Figure 11C); LNCaP-luc cells
showed dramatic declines in SREBP-1c, FAS and ACC1, but marked elevations in
SCAP. HP-LNCaP cells had increased expression of SREBP-1a, SCAP, and ACC1
compared to the parent line (Figure 11C). Though passage altered the expression of
several mitochondrial genes in HP-LNCaP cells, the LNCaP-luc cells were more similar
to the low passage parental line.
The LNCaP derived lines showed different patterns in terms of mitochondrial
enzyme properties. LNCaP-luc cells showed approximately 50% reductions in both CS
and COX (Figure 10C), which is similar to what was seen with PC3 and DU145 (Figure
10). In contrast, HP-LNCaP cells possessed CS and COX activities that were near that of
the parental line.
50
Figure 10. Metabolism of HP-LNCaP and LNCaP-luc cells
A) O2 consumption, and B) lactate production. Bars are means (+SEM) with different
letters denoting significantly different values within each cell line according to ANOVA
and Tukey tests (n=6). C) Oxidative enzyme activity. Bars are means (+SEM) with
different letters denoting significantly different values for each enzyme assayed
according to ANOVA and Tukey tests (n=5). All enzyme levels are displayed relative to
LNCaP means: CS 0.33 ± 0.04, COX 0.02 ± 0.004 enzyme units (nmol/min/mg protein).
51
Figure 11. Transcript levels in HPLNCaP and LNCaP-luc cells
A) mRNA levels of hypoxia response
genes, B) mRNA levels of
mitochondrial genes and
transcriptional regulators, and C)
SREBP axis genes. All transcript
levels displayed relative to LNCaP
transcript levels. Bars are means
(+SEM) with different letters denoting
significantly different values between
cell lines according to non-parametric
Wilcoxon tests (n=6). See
Supplementary Table 2 for raw cycle
thresholds relative to TBP.
52
2.3.4 Metabolic gene expression of LNCaP cells in vivo
To investigate whether mRNA levels differ between cells grown in vitro and in
vivo, we analyzed expression profiles of HP-LNCaP and LNCaP-luc cells grown as
xenografts in SCID mice.
HP-LNCaP xenografts had significantly greater mRNA levels of HIF-1α and
VEGF65 than cells grown in vivo; however, the only notable change in the glycolytic
genes was a significant decrease in ALDOa transcript abundance in vivo (Figure 12A). In
comparison to cells grown in vitro, HP-LNCaP xenografts showed an increase in the
levels of PGC-1α (7-fold), COXIV (3-fold) and COXI (5-fold) mRNA levels (Figure
12B). In terms of lipogenesis, xenografts displayed a significantly lower abundance of
SREBP-1c and ACC1 transcripts, with a 5-fold increase in SCAP mRNA levels (Figure
12C).
53
Figure 12. Transcript levels in HPLNCaP cells and xenografts
A) mRNA levels of hypoxia response
genes, B) mRNA levels of mitochondrial
genes and transcriptional regulators, and
C) mRNA levels of SREBP axis genes.
Bars represent means (+SEM) mRNA
levels displayed relative to mRNA levels
in HP-LNCaP cells. * Indicates
significant difference according to nonparametric Wilcoxon tests (n=6). See
Supplementary Table 3 for raw cycle
thresholds relative to TBP.
54
LNCaP-luc cells were grown as xenografts in kidney and leg muscle (Figure 13).
There were few notable differences in these cells when grown in the different contexts.
There was some evidence of a hypoxic response in both LNCaP-luc xenografts in vivo,
with elevations in one VEGF isoform (muscle only) and selected glycolytic enzymes
(HKII, PKm) (Figure 13A). Both kidney and muscle xenografts showed an increase in
COXI (Figure 13B).
55
Figure 13. Transcript levels in
LNCaP-luc cells and xenografts
A) mRNA levels of hypoxia related
genes, and B) mRNA levels in
mitochondrial genes and
transcriptional regulators. Bars
represent mean (+SEM) mRNA
levels displayed relative to
transcript levels in LNCaP-luc cells.
* Indicates significant difference
according to non-parametric
Wilcoxon tests (n=6). See
Supplementary Table 4 for raw
cycle thresholds relative to TBP.
56
2.3.5 SREBP axis gene expression is not hypoxia sensitive
Cells were exposed to hypoxia, anoxia and azide for 12 hours to determine the
sensitivity of the SREBP axis to low oxygen using real-time PCR (Figure 14). In each
cell line, the levels of VEGF mRNA (a HIF-1 target) significantly increased following
treatment with anoxia. Anoxic treatment of LNCaP and PC3 cells (though not DU145
cells) also increased the abundance of HKII transcripts, another HIF-1 target. Less of a
response was seen with hypoxia, and azide had no effect. In contrast, anoxia had no effect
on the mRNA levels of FAS, ACC1, or SCAP in any cell line (Figures 14 C, D, G). PC3
cells, but not the other lines, showed a significant increase in the levels of SREBP-1a and
SREBP-1c mRNA (Figures 14 E, F).
57
Figure 14. mRNA levels in LNCaP, DU145 and
PC3 cells following exposure to hypoxia, anoxia, or
azide
mRNA levels of the hypoxia response genes VEGF
(A) and HKII (B), the SREBP downstream targets
FAS (C) and ACC1 (D), the transcription factors
SREBP-1a (E) and SREBP-1c (F), and the SREBP
cleavage activating protein (G). Bars represent means
(+SEM) mRNA levels displayed relative to control
transcript levels. * Indicates significant difference
within each cell line according to non-parametric
Wilcoxon tests (n=5). See Supplementary Table 5 for
raw cycle thresholds relative to TBP.
58
2.3.6 Fatty acid levels are not influenced by low oxygen in these prostate
cancer models
Cells were exposed to normoxia, hypoxia, and anoxia for a 48-hour period to
determine the influence of low oxygen on their steady-state triglyceride levels (Figure
15). No significant differences were noted between the control samples and the low
oxygen treatments for any of the cell lines (Figure 15).
Figure 15. Triglyceride content
(ug TG/mg protein) of LNCaP,
DU145, and PC3 cells following
exposure to normoxia, hypoxia,
and anoxia.
Bars represents means (+SEM) of all
cell lines relative to the control and *
indicates significance according to
non-parametric Wilcoxon tests
(n=6).
2.4 Discussion
Although there is a common perception that cancer cells support their energetic
demands through aerobic glycolysis, some cancer cells rely heavily on OXPHOS [15]. In
this study we investigated the most widely used of prostate cancer cell lines: LNCaP,
DU145, and PC3. The three classical prostate cancer cell lines show dramatic differences
in their metabolic phenotype, manifested as differences in aerobic glycolysis and
59
oxidative metabolism. LNCaP cells are highly oxidative in contrast to both DU145 and
PC3 lines. That is, LNCaP cells have higher rates of oxygen consumption and lower rates
of lactate production when in comparison to the other lines. Such differences in the
relative importance of oxidative metabolism would be expected to influence the efficacy
of drugs with metabolic targets. For example, DCA has been proven successful at
reversing the glycolytic phenotype of non-small-cell lung cancer by simultaneously
increasing the oxidation of glucose in the mitochondria, and decreasing the rates of
lactate production and fatty acid oxidation [28]. Though LNCaP cells did increase their
rate of respiration when treated with DCA, the other two cell lines were unresponsive to
the treatment. Thus, our studies suggest that the small molecule DCA exerts its greatest
effects on metabolism in cells with an oxidative metabolic poise.
2.4.1 Metabolic enzyme activities and gene expression
In the cancer lines we studied, there was a general maintenance in enzyme
stoichiometries within the glycolytic pathway and a general similarity between the lines.
Despite the similarities in enzyme activites, there were intriguing differences in other
aspects of the glycolytic phenotype. There was a loss of stoichiometry in the relationships
between transcript abundance and enzyme activities; PC3 cells showed higher mRNA
levels for all glycolytic enzymes studied. This is likely attributed to differential posttranscriptional regulation of specific mRNAs encoded by glycolytic genes. As well, both
PC3 and DU145 cells showed circa 10-fold higher HIF-1a mRNA levels. HIF-1a levels
are normally regulated posttranslationally (Huang et al. 1996), and hypoxia generally
60
doesn’t affect HIF-1a transcription (but see Zhong et al. 1998). Nonetheless, both PC3
and DU145 cells showed parallel elevations in mRNA of both HIF-1a and VEGF
isoforms (HIF-1 targets, Semenza et al. 1996) (Figure 6A).
Despite this apparent
“hypoxic response” in DU145 and PC3, only PC3 cells showed elevated levels of
mRNAs encoded by glycolytic genes (Figure 6A). It was also noteworthy that, despite
similar glycolytic enzyme levels, the cells demonstrated differences in glycolytic rate.
The glycolytic phenotype of PC3 and DU145 cells is most likely due to
mitochondrial insufficiency, due either to lower mitochondrial content or mitochondrial
dysfunction. In this study, the two glycolytic lines had a 50-65% reduction in the
activities of two mitochondrial enzymes (COX and CS) compared to LNCaP cells. This
pattern is consistent with previous studies in other cancer lines (Miralpeix et al. 1990;
Matoba et al. 2006). However, the two glycolytic lines show no reduction in levels of
mRNAs encoding mitochondrial enzymes. In fact, relative to LNCaP cells, DU145 cells
show 5-fold greater expression of CS and COX subunits. In both glycolytic lines, the
mRNA levels of important regulators of mitochondrial gene transcription (PGC-1α and
NRF-2), are dramatically elevated. With an elevation in the mRNA levels of stimulators
of mitochondrial biogenesis, the most parsimonious explanation for depressed
mitochondrial enzyme levels is likely a disruption of mitochondrial biogenesis, rather
than a regulated decrease in the content of otherwise normal mitochondria. Each of the
cell lines possesses similar cardiolipin levels, suggesting that the content of mitochondria
is comparable between lines. Thus, the reduced respiration rates and lower mitochondrial
61
enzyme activities are consistent with less effective mitochondria, rather than fewer
mitochondria. The PC3 and DU145 cells compensate for the energetic shortfalls by
stimulating glycolysis, likely achieved not through elevated glycolytic enzyme levels but
through the normal allosteric and covalent regulators of glycolysis. Increased glycolytic
flux is likely a response to this dysfunction, and achieved not through elevated glycolytic
enzyme levels but through the normal allosteric and covalent regulators of glycolysis.
The exact origins of the mitochondrial defects in the PC3 and DU145 lines remain
unclear. Although the PC3 and DU145 cells possessed lower mitochondrial enzyme
levels, they showed an apparent stimulation of genes associated with mitochondrial
enzymes and their regulators. This would appear to represent a futile effort to stimulate
mitochondrial biogenesis, perhaps in response to energetic shortfalls. The resulting state
is fairly characterized as a mitochondrial lesion, where defects in one or more critical
enzymes lead to an inability to sustain mitochondrial biogenesis. In other situations,
mitochondrial lesions are attributed to defects in COX (Leary et al. 2004), including
defects in processing of the heme and copper groups that are essential for the holoenzyme
(Capaldi 1990). Matoba et al. (2006) proposed that a loss of p53 activity might reduce
COX levels by inhibiting the expression of the SCO2 gene, which encodes a protein that
participates in copper delivery to the mitochondria. Though we have demonstrated
DU145 and PC3 cells have reduced COX enzyme activities, and these lines have
previously been characterized to lack functional p53 (Carroll et al. 1993), their SCO2
expression is no different from that of the LNCaP cells. Thus, the reductions in
62
mitochondrial enzymes in PC3 and DU145 cells cannot be explained by reduced
transcription of (i) the respiratory transcriptional regulators of biogenesis (PGC-1,
NRF2), (ii) the genes encoding the oxidative enzymes or (iii) the SCO2 assembly factor.
2.4.2 The effects of passage and selection
While we have established that aerobic glycolysis is an important characteristic of
prostate cancer cells, the evolution of this phenotype has not been sufficiently described.
Somatic evolution during carcinogenesis usually involves the combination of genetic
events such as proto-oncogene gain-of-function, loss of tumour suppressor function, and
environmental selection (Vogelstein and Kinzer 2004; Smalley et al. 2005). In order for a
particular phenotype to persist evolutionarily, it must confer a selective advantage. If the
glycolytic phenotype serves to supply ATP when oxidative metabolism is lacking, it is
peculiar that when oxygen is abundant that the phenotype is maintained (Costello and
Franklin 2005). We have demonstrated that with prolonged passage in normoxia, LNCaP
cells evolve to become more glycolytic as demonstrated by their decreased rate of oxygen
consumption and increased rate of lactate production. Additionally, when LNCaP cells
are clonally selected, an even more dramatic transformation occurs. Bordeau-Heller and
Oberley (2007) demonstrated that following prolonged culture in hypoxic conditions that
DU145 cells significantly reduced their rate of oxygen consumption and increased HIF1α protein levels. Interestingly, in LNCaP cells cultured under normoxia, a similar
metabolic transition to aerobic glycolysis occurred. The exact mechanism by which this
metabolic transformation occurs remains unclear. Alternatively, it is possible that a loss
63
of p53 activity occurs in LNCaP cells leading to selection for a more resistant cancer that
is able to withstand reduced oxygen levels and acidity, and resist apoptotic signals.
Based on the collective results, it appears that the aerobic glycolytic phenotype
displayed in select prostate lines and with passage is a response to mitochondrial
shortfalls, the origin of which remain unknown.
2.4.3 In vivo and in vitro differences in gene expression
Cell culture models are commonly used in cancer biology, however,
relatively little is known about how metabolic gene expression in cultured cells relates to
gene expression in vivo. In vitro, there is a tight relationship between gene expression,
growth factor signaling, proliferation and treatment efficacy (Rice et al 1986; Sutherland
1988; Moscow and Cowen 1988). In vivo oncogenesis is much more complex for it
requires the development of a new vascular network to allow individual cells to be
perfused with oxygen and nutrients from the circulation. The microcirculation and degree
of vascularization of a particular tumour influences the degree of hypoxia and acidosis in
its microenvironment (Kallinowski et al. 1989). Therefore gene expression in vivo may
be governed by the interplay between the endogenous metabolic phenotype of a cell and
the environment in which it is found. To explore this question, we studied two derived
lines of LNCaP cells to assess how expression patterns were affected by growth in a more
complex tumour environment.
While HP-LNCaP tumour xenografts were from the
kidney; LNCaP-luc cells were grown as xenografts in both kidney (highly perfused) and
leg muscle (less vascularized).
64
Though there were differences between LNCaP derived lines, there were no
parallels between the xenografts of the high passage and luciferase-transfected cells. HPLNCaP xenografts had a moderate increase in expression of HIF-1α and VEGF65, and a
much greater stimulation of mitochondrial gene expression. Contrarily, the expression
patterns of glycolytic and mitochondrial genes were nearly indistinguishable in LNCaPluc cells grown in culture versus in xenografts. These results are consistent with the
findings of Kallinowski et al. (1989), who demonstrated that the balance between
glycolysis and oxidative phosphorylation in xenografts is determined independently of
oxygen and glucose delivery. We observed that, while the expression of both VEGF
isoforms significantly differed between the muscle and leg LNCaP-luc xenografts, no
differences were observed in any of the other glycolytic genes, which are also hypoxia
responsive (Semenza et al. 1996). On the SREBP axis, muscle LNCaP-luc xenografts had
a slightly higher expression of both SREBP-2 and FAS compared to both cultured cells
and kidney xenografts. Conversely, SREBP-1c and ACC1 transcript levels were
significantly lower in the HP-LNCaP xenografts. These studies suggest that individual
cells are selected for a response according to their genotype rather than the surrounding
tissue in which they are found. Furthermore, while the microenvironment is of
importance in tumours growing in patients, it is not solely responsible (in the models
examined here) for the responses seen in vivo.
65
2.4.4 SREBP axis
In addition to the Warburg effect, many cancer cells have been shown to upregulate the de novo fatty acid synthesis pathway. The relationship between hypoxia and
lipid metabolism has yet to be completely resolved. Glycolytic pyruvate can be either
converted to lactate in the cytosol, or imported into mitochondria to form citrate, which is
exported to the cytosol and cleaved to produce the acetyl-CoA used in fatty acid synthesis
(reviewed by Swinnen et al. 2006). Both lactate production and fatty acid synthesis
require reducing equivalents and thus may both serve to maintain redox balance in cancer
cells when oxygen is limiting. Whether the SREBP axis is induced to serve this purpose,
or merely to provide membrane phospholipids during proliferation is unknown. Our gene
expression studies examining the sensitivity of the SREBP axis to low oxygen suggest
that oxygen does not influence lipogenic gene transcription.
Other regulatory cascades have been shown to affect lipid metabolic enzymes.
Growth factors such as epidermal growth factor (EGF) and keratinocyte growth factor
(KGF, also known as fibroblast growth factor 7, FGF-7) can stimulate lipogenic gene
transcription. Insulin can also stimulate lipid metabolic enzymes via the PI3K cascade
(Swinnen et al. 2000; Chang et al. 2005). Though hypoxia may not induce the expression
of the genes associated with fatty acid synthesis (ACC1, FAS), or affect the transcription
of the regulators of these genes (SREBP, SCAP) this does not preclude a role for fat
synthesis in hypoxic energy metabolism.
66
While many studies have characterized lipogenic gene expression in cancer cells,
few have actually quantified the triglycerides the cells produce. Since our gene
expression results suggest that the relationship between hypoxia and fatty acid synthesis
do not occur on a transcriptional level, we proceeded to quantify the triglycerides
following treatment. We found that exposure to low oxygen had no significant impact on
the production of triglycerides. Furthermore, we found no differences in baseline
triglyceride production between the cells lines. Therefore we conclude that de novo fatty
acid synthesis is not related to low oxygen, and furthermore, that the metabolic
phenotype of a cell in terms of glycolytic and oxidative balance does not dictate its
degree of fatty acid production.
Collectively, our results illustrate the plasticity of the metabolic phenotype and
highlight the importance of considering how a cell’s endogenous metabolism can
influence its ability to survive and progress in hypoxic conditions. These data support the
model that a loss of mitochondrial function is responsible for aerobic glycolysis in these
prostate cancer cells. Further insight into the relationship between hypoxia and fatty acid
synthesis is required to determine whether or not this pathway is involved in the cell’s
response to hypoxia.
67
Chapter 3: General Discussion
3.1 The metabolic phenotype of prostate cancer cells
Cells require energy in the form of ATP to carry out vital anabolic processes
involved in growth and repair. Such ATP is produced during the metabolic processes of
glycolysis, the TCA cycle, and oxidative phosphorylation. Glycolysis occurs in the
cytoplasm, and is functional even in the absence of oxygen; however, it only yields two
ATP and two NADHs per glucose molecule. In contrast, oxidative phosphorylation
occurs in the mitochondrial electron transport chain, requires oxygen, and produces as
many as 38-39 ATP per molecule of glucose. Due to its high efficiency in terms of ATP
produced per glucose, it is evident that during aerobic conditions this pathway would be
preferred.
When oxygen levels are reduced, cells produce most of their energy using
glycolysis, since aerobic pathways cannot be used. Under such circumstances, pyruvate is
converted to lactate by LDH in the cytoplasm, which is eventually shuttled out of the cell.
For example, glycolytic metabolism is used in white muscle during burst exercise, when
the cell requires ATP at a higher rate than it can get from mitochondrial pathways.
Similarly, many cancer cells have been shown to have an unusual metabolic phenotype
where they derive most of their cellular energy from glycolysis, even in the presence of
oxygen. This is referred to aerobic glycolysis, and while its occurrence may be explained
in many ways, it is still widely debated and largely misunderstood.
68
The first objective of my thesis was to determine the metabolic phenotype of
prostate cancer cells. I chose to examine three prostate cancer cell lines in particular:
LNCaP, DU145, and PC3 cells. These cells are widely used models for prostate cancer,
but how they differ metabolically has yet to be investigated. It is often generalized that all
cancer cells exhibit the phenotype of aerobic glycolysis; however, following examination,
it seems that some cancer cells do not. For example, certain types of lung cancer cells
(Weinhouse 1956), bone sarcomas (Zu and Guppy 2004), and uterine cancer cells
(Kallinowski et al. 1989) are documented as being predominantly oxidative. Therefore, I
set out initially to describe these prostate cancer cell lines metabolically, and to determine
whether they predominantly use glycolytic or oxidative metabolism for ATP production.
I demonstrated that LNCaP cells are much more oxidative than the DU145 and
PC3 cells. Comparatively, LNCaP cells had a greater rate of oxygen consumption and
lower rate of lactate production rate than the other cell types. Furthermore, when exposed
to acidity and hypoxia in their environment, LNCaP cells responded much more
dramatically in terms of their ATP levels, whereas the DU145 and PC3 cells were already
seemingly hypoxic. When I investigated the activities of the mitochondrial enzymes CS
and COX, I discovered that these were reduced significantly in the DU145 and PC3 cells,
while their mRNA levels were at par with those of the LNCaP cells. This was the first
clue that respiratory impairment may be responsible for the aerobic glycolytic phenotype
of the glycolytic lines. Taken alone, this could be due to: 1) defects in mitochondrial
biogenesis, or 2) mitochondrial dysfunction. Following quantification of mitochondrial
69
content with cardiolipin staining, I discovered no differences between the cell lines,
suggesting that mitochondrial dysfunction was likely the cause of the glycolytic
phenotype observed in the DU145 and PC3 cells. When I examined this further using
DCA and DNP, I discovered that while respiration could be stimulated in the LNCaP
cells, this was not the case with the glycolytic lines. Therefore, mitochondrial dysfunction
seems to play a role in the glycolytic phenotype of the DU145 and PC3 cells, which
supports Warburg’s hypothesis that aerobic glycolysis results due to respiratory chain
deficiencies.
While mitochondrial dysfunction clearly plays a role in aerobic glycolysis in the
DU145 and PC3 cells, the mechanistic basis of this dysfunction remains unknown. There
are some major molecular differences between these prostate cancer cell lines that I
believe to be of critical importance. Firstly, LNCaP cells express the androgen receptor
and are an excellent model for studying androgen-dependent earlier stages of prostate
cancer. Contrarily, DU145 and PC3 cells are androgen-independent cells and represent a
more advanced stage of the disease. LNCaP cells secrete PSA, and PC3 and DU145 cells
do not. In addition, LNCaP cells have functional p53, whereas the glycolytic lines do not.
Thus, the genetics of these cancer models is almost sufficient to explain their metabolic
differences, as the androgen receptor and p53 are both part of many complex signaling
pathways within the cell. Furthermore, these results emphasize the importance of
studying multiple cell types during cancer research, because different cells of a tumour
may have metabolic differences that might influence the efficacy of treatment. This holds
70
true for all types of cancers because the mutations the cells have incurred will most
definitely influence treatment efficacy. It is therefore of extreme importance to not
generalize results found in one cell type to the over phenotype of the cancer type.
3.2 The influence of passage and selection on prostate cancer cells and
xenografts
Somatic evolution during carcinogenesis involves a series of mutations in protooncogenes and tumour suppressor genes, which ultimately lead to a survival advantage
for the cell. Additionally, environmental selection plays a role in the integration if these
genetic alterations into the overall phenotype of the cell. While the metabolic phenotype
differs between prostate cancer cell lines, little is known about how this phenotype
evolves with passage and clonal selection. The LNCaP line is commonly used in studies
examining the early, androgen-dependent stages of prostate cancer. Since aerobic
glycolysis is associated with increased aggressiveness and therapeutic resistance, it is
quite possible that over time LNCaP cells might evolve a glycolytic phenotype.
I investigated the plasticity of the metabolic phenotype using the LNCaP-derived
lines, HP-LNCaP and LNCaP-luc. In terms of oxygen consumption and lactate
production, I found the LNCaP-derived lines to be more glycolytic than the parent line.
Higher levels of HIF-1α and VEGF mRNA were seen in the LNCaP-luc and HP-LNCaP
cells, as well as an increase in expression of certain glycolytic genes. CS enzyme levels
were reduced by 60% in the clonally selected line; however, no significant changes were
noted in oxidative enzyme levels in the HP-LNCaP cells. Mitochondrial content in the
71
LNCaP-derived lines was not significantly different that in any of the other lines
evaluated, ruling out problems with the mitochondrial biogenesis. Therefore,
mitochondrial dysfunction may explain why, under normoxic conditions, the LNCaP
cells evolved a glycolytic phenotype. This suggests that the phenotype of aerobic
glycolysis can be selected for with time in culture and clonal selection from a more
oxidative phenotype, and illustrates the importance of aerobic glycolysis in cancer
development and progression. Additionally, we cannot overlook the fact that the passage
history of most cell lines used in research is somewhat of a mystery, which makes it
difficult to determine whether results are significant or merely due to culture artifacts.
The ability of the LNCaP cells to become more glycolytic with passage and
selection in culture is extremely interesting for it may represent prostate cancer
progression to androgen-independence. This is interesting because the derived lines
became more metabolically similar to the DU145 and PC3 cells lines than they are to
their parent line. While LNCaP cells are androgen sensitive, DU145 and PC3 cells
represent models for androgen-independent prostate cancer. Therefore it is quite possible
that the metabolic transformation I observed was coordinated with progression to
androgen independence. Thus, genetic alterations resulting in the emergence of androgen
independence, such as activation of the Akt pathway, or loss of p53 function may be
sufficient to explain the metabolic change that occurred in the LNCaP cells. Overall,
clonal selection seemed to have a greater effect on metabolism than did passage in
culture; however, both resulted in the aerobic glycolytic phenotype. Interestingly
72
however, these two lines were chosen purely for opportunistic reasons and a replication
of the same experiment with alternative available lines would most likely result in
derived lines with novel properties.
In addition to cell culture models, xenograft models are commonly used in cancer
research for they are more representative of cancer as it occurs in vivo. Little is known
about how gene expression differs between culture and xenograft models. Xenografts are
much more complex because the degree of vascularization dictates the cell’s ability to
access oxygen and nutrients from the circulation. In contrast, these same constraints do
not exist in cell culture models. It has even been argued that in cell culture systems
oxygen levels are in fact well above normal levels in the body, and that this hyperoxia
might have its own consequences for metabolism and/or selection. I therefore
investigated metabolic gene expression in the LNCaP-derived cell lines as well as in their
xenograft counterparts in the kidney and muscle to explore the plasticity of the response
of a given cell line. While I noted a greater hypoxia response (in terms of HIF-1α and
VEGF expression) in the HP-LNCaP xenografts when compared to the cells, PGC-1α,
COXIV and COXI mRNA levels were also significantly increased. This may in part be
due to intermittent hypoxia exposure, where the cells of the tumour are exposed to
periods of hypoxia followed by oxygenation, and attempt to compensate by increasing
the expression of the transcriptional regulators of oxidative metabolism. In the LNCaPluc cells, some glycolytic genes had increased mRNA levels in vivo; however, HIF-1α
mRNA levels were actually lower than they were in the LNCaP-luc cells. I therefore
73
concluded from this that there are few parallels between cell culture and xenograft
models of prostate cancer, and each of these models is likely to differ from that of a
human tumour. Similarly, gene expression would likely differ between different tumours,
because the environment in which the tumour is found plays a significant role in the
signaling pathways that occur at a cellular level.
3.3 The influence of hypoxia on fatty acid synthesis
Fatty acid synthase, the final enzyme in the pathway of de novo fatty acid
synthesis is upregulated at the mRNA and protein level in prostate cancer cells, and the
reasons for this remain debated (Shurbaji et al. 1992). This pathway ends with the
production of the fatty acid palmitate which is subsequently used in the formation of
membrane phospholipids or TGs for storage. During periods of hypoxia, this pathway
may be upregulated in order to maintain redox balance to compensate for the reduction in
respiratory chain activity (Hochachka et al. 2002). Much more redox power is required
from the pathway of fatty acid synthesis in comparison to lactate production, so perhaps
it represents the means by which the cell oxidizes the NADHs produced from glycolysis
and the TCA cycle. I set out to determine the relationship between hypoxia and fatty acid
synthesis, more specifically to determine if the upregulation in this pathway is directly a
result of oxygen deprivation.
Firstly, I evaluated whether hypoxia, anoxia, or the chemical hypoxia mimetic
azide influence the mRNA levels of genes in the SREBP axis. I found no significant
upregulation at the mRNA level, suggesting that the pathway of de novo fatty acid
74
synthesis is not hypoxia-sensitive. However, in the PC3 cells, there was an upregulation
of SREBP-1a and SREBP-1c transcripts after 12 hours of treatment, which suggested that
my treatment times were perhaps not long enough to see a response. Given the half-life of
most mRNA (at least those lacking regulatory sequences in the upstream and downstream
untranslated region) is around 4h. Thus it is likely that any response seen after 12 h is not
a primary response to low oxygen but rather a secondary response to cellular changes
made during hypoxia. However, when I quantified the levels of TGs in the cells
following 48 hours of hypoxic or anoxic exposure, I reached a similar conclusion; low
oxygen had no influence on the production of TGs in any of the cell lines. In addition, no
significant differences were found in the quantity of TGs between cell lines under
normoxia despite their very different metabolic phenotypes. Therefore, while fatty acid
synthesis is an important part of the metabolism of a cell, it is not directly involved in the
hypoxia response or aerobic glycolysis.
The results from my thesis can be summarized briefly. 1) LNCaP cells are
oxidative, whereas DU145 and PC3 cells exhibit aerobic glycolysis. 2) Mitochondrial
dysfunction plays a role in the glycolytic phenotype of DU145 and PC3 cells. 3) The
phenotype of aerobic glycolysis can be selected for with passage in culture and clonal
selection. 4) Gene expression between culture and xenograft models is highly
unpredictable. 5) There is no direct relationship between hypoxia and fatty acid synthesis
in prostate cancer cells.
75
These results bring about endless questions that require experimental
investigation. While I have shown that mitochondrial dysfunction is involved in the
aerobic glycolytic phenotype of DU145 and PC3 cells, it is still unknown whether
mitochondrial dysfunction is the cause of aerobic glycolysis, or whether it occurs as a
consequence or adaptation to this altered metabolism. Perhaps these results can all be
explained by the fact that LNCaP cells are androgen-sensitive and DU145 and PC3 cells
are not. The differences in cell lines I observed underscore the importance of using these
important cell models of cancer complimentarily and with caution when translating the
conclusions to treatment of prostate (or other) cancers. Prostate cancer is a complex
disease, for initially it requires hormones to grow and survive, but eventually advances to
a more aggressive androgen-independent cancer. With this progression to androgenindependence, a patient’s chances of surviving are dramatically reduced (Petrylak et al.
2004; Shamash et al. 2005). Though it would be worthwhile to study the androgen
dependence of metabolic patterns of these models, such studies were beyond the scope of
my thesis. An immense number of signaling pathways are involved in integrating
environmental signals with the cell’s metabolism and these play a role in progression to
an androgen-independent metastatic cancer. These pathways must be delineated and the
detailed relationships between the HIF-1, p53, Akt, and AR pathways must be fully
understood in order to fully understand aerobic glycolysis. Nonetheless, studies
evaluating the metabolic properties of cancer cells are of the utmost importance, and must
be considered when targeting aspects of metabolism for cancer treatment.
76
Literature Cited
Acs, G., Xu, X., Chu, C., Acs, P., and Verma, A. (2004). Prognostic
significance of erythropoietin expression in human endometrial carcinoma. Cancer. 100,
2376-2386.
Akakura, N., Kobayashi, M., Horiuchi, I., Suzuki, A., Wang, J., Chen, J.,
Niizeki, H., Kawamura, K.-I., Hosokawa, M. and Asaka, M. (2001). Constitutive
Expression of Hypoxia-inducible Factor-1 Renders Pancreatic Cancer Cells Resistant
to Apoptosis Induced by Hypoxia and Nutrient Deprivation. Cancer Res. 61, 6548-6554.
Bacus, S.S., Altomare, D.A., Lyass, L., Chin, D.M., Farrell, M.P., Gurova,
K., Gudkov, A., and Testa, J.R. (2002). AKT2 is frequently upregulated in HER-2/neupositive breast cancers and may contribute to tumor aggressiveness by enhancing cell
survival. Oncogene. 21, 3532-3540.
Bellacosa, A., de Feo, D., Godwin, A.K., Bell, D.W., Cheng, J.Q., Altomare,
D.A., Wan, M., Dubeau, L., Scambia, G., Masciullo, V., Ferrandina, G., Benedetti
Panici, P., Mancuso, S., Neri, G., and Testa, J.R. (1995). Molecular alterations of the
AKT2 oncogene in ovarian and breast carcinomas. 64, 280-285.
Bensaad, K., Tsuruta, A., Selak, M.A., Calvo Vidal, M.N., Nakano, K.,
Bartrons, R., Gottlieb, E., and Vousden, K.H. (2006). TIGAR, a p53-inducible
regulator of glycolysis and apoptosis. Cell. 126, 107-120.
Beutner, G., Ruck, A., Riede, B., and Brdiczka, D. (1998). Complexes
between porin, hexokinase, mitochondrial creatine kinase and adenylate translocator
display properties of the permeability transition pore: Implication for regulation of
permeability transition by the kinases. BBA. 1368, 7-18.
Bonnet, S., Archer, SL., Allalunis-Turner, J., Haromy, A., Beaulieu, C.,
Thompson, R., Lee, C.T., Lopaschuk, G.D., Puttagunta, L., Bonnet, S., Harry, G.,
Hashimoto, K., Porter, C.J., Andrade, M.A., Thebaud, B., and Michelakis, E.D.
(2007). A mitochondrial-K+ axis is suppressed in cancer and its normalization promotes
apoptosis and inhibits cancer growth. Cancer Cell. 11, 37-51.
Bourdeau-Heller, J., and Oberley, T.D. (2007). Prostate Carcinoma Cells
Selected by Long-term Exposure to Reduced Oxygen Tension Show Remarkable
Biochemical Plasticity Via Modulation of Superoxide, HIF-1a Levels, and Energy
Metabolism. J Cell Physiol. 212, 744-752
Brown, M.S., and Goldstein, J.L. (1997). The SREBP pathway: Regulation of
cholesterol metabolism by proteolysis of a membrane-bound transcription factor. Cell.
77
89, 331-340.
Brusselmans, K., De Schrijver, E.D., Verhoeven, G., and Swinnen, J.V.
(2005). RNA interference –mediated silencing of the acetyl-CoA-carboxylase-a gene
induces growth inhibition and apoptosis of prostate cancer cells. Cancer Res. 65, 67196725.
Cantley, L.C. (2002). The phosphoinositide 3-kinase pathway. Science. 296,
1655–1657.
Capaldi, R.A. (1990). Structure and assembly of cytochrome c oxidase. Arch
Biochem Biophys. 280, 252-262.
Carmeliet, P., Dor, Y., Herbert, J.-M., Fukumura, D., Brusselmans, K.,
Dewerchin, M., Neeman, M., Bono, F., Abramovitch, R., Maxwell, P., Koch, C.J.,
Ratcliffe, P., Moons, L., Jain, R.K., Collen, D., and Keshet, E. (1998). Role of HIF1α in hypoxia-mediated apoptosis, cell proliferation and tumour angiogenesis. Nature
1998; 394: 485-490.
Carroll, A.G., Voeller, H.J., Sugars, L., and Gelmann, E.P. (1993). P53
oncogene mutations in three human prostate cancer cell lines. Prostate. 23, 123-134.
Chang, Y., Wang, J., Lu, X., Thewke, D.P., and Mason, R.J. (2005) KGF
induces lipogenic genes through a PI3K and JNK/SREBP-1 pathway in H292cells. J
Lipid Res. 46, 2624-2635.
Chen, C.D., Welsbie, D.S., Tran, C., Baek, S.H., Chen, R., Vessella, R.,
Rosenfeld, M.G., and Sawyers, C.L. (2004). Molecular determinants of resistance to
antiandrogen therapy. Nat Med. 10, 33–39.
Chung, L.W.K, Li, W., Gleave, M.E., Hsieh, J.T., Wu, H.-C., Sikes, R.A.,
Zhau, H.E., Bandyk, M.G., Logothetis, C.J., Rubin, J.S., and von Eschenbach A.C.
(1992) Human prostate cancer model: roles of growth factors and extracellular matrices.
J Cell Biochem. Suppl, 99-105.
Clarke, A.E., West, K., Streicher, S., and Dennis, P.A. (2002). Constitutive
and inducile Akt activity promotes resistance to chemotherapy, trastuzamab, and
tamoxifen in breast cancer cells. Mol Cancer Ther. 1, 707-717.
Coffey, R.N.T., Morrissey, C., Taylor, C.T., Fitzpatrick, J.M., and Watson,
R.W.G. (2005). Resistance to caspase-dependent, hypoxia-induced apoptosis is not
hypoxia-inducible factor-1 alpha mediated in prostate carcinoma cells. Cancer. 103,
78
1363-1374.
Costello, L., and Franklin, R. (2005). Why do tumour cells glycolyse? From
glycolysis through citrate to lipogenesis. Mol Cell Biochem. 280, 1-8.
Culig, Z., Hobisch, A., Cronauer, M.V., Radmayr, C., Trapman, J.,
Hittmair, A., Bartsch, G., and Klocker, H. (1994). Androgen receptor activation in
prostatic tumor cell lines by insulinlike growth factor-I, keratinocyte growth factor, and
epidermal growth factor. Cancer Res. 54,5474–5478.
Dachs, G.U., and Tozer, G.M. (2000). Hypoxia modulated gene expression:
angiogenesis, metastasis and therapeutic exploitation. 36, 1649-1660.
Debes, J.D., and Tindall, D.J. (2004). Mechanisms of androgen-refractory
prostate cancer. NEJM. 351, 1488-1490.
Dement, G.A., Maloney, S.C., and Reeves, R. (2007). Nuclear HMGA1
nonhistone chromatin proteins directly influence mitochondrial transcription,
maintenance, and function. Exp Cell Res. 2007. 313, 77-87.
Eberlé, D., Hegarty, B., Bossard, P., Ferré, P., and Foufelle, F. (2004).
SREBP transcription factors: master regulators of lipid homeostasis. Biochimie. 86, 839848.
Ebert, B.L., Firth, J.D., and Ratcliffe, P.J. (1995). Hypoxia and mitochondrial
inhibitors regulate expression of glucose transporter-1 via distinct cis-acting sequences. J
Biol Chem. 270, 29083–29089.
Elstrom, R.L., Bauer, D.E., Buzzai, M., Karnauskas, R., Harris, M.H., Plas,
D.R., Zhuang, H., Cinalli, R.M., Alavi, A., Rudin, C.M., and Thompson, C.B.
(2004). Akt stimulates aerobic glycoysis in cancer cells. Cancer Res. 64, 3892-3899.
Epstein, A.C., Gleadle, J.M., McNeill, L.A., Hewitson, K.S., O’Rourke, J.,
Mole, D.R., Mukherji, M., Metzen, E., Wilson, M.I., Dhanda, A., Tian, Y.M.,
Masson, N., Hamilton, D.L., Jaakkola, P., Barstead, R., Hodgkin, J., Maxwell, P.H.,
Pugh, C.W., Schofield, C.J., and Ratcliffe, P.J. (2001). C. elegans EGL-9 and
mammalian homologs define a family of dioxygenases that regulate HIF by prolyl
hydroxylation. Cell. 107, 1-3.
Ettinger, S.L., Sobel, R., Whitmore, T.G., Akbari, M., Bradley, D.R., Gleave,
M.E., and Nelson, C.C. (2004). Dysregulation of sterol response element-binding
proteins and downstream effectors in prostate cancer during progression to androgen
79
independence. Cancer Res. 64, 2212-2221.
Feilchenfeldt, J., Brundler, M.A., Soravia, C., Totsch, M., and Meier, C.A.
(2004). Peroxisome proliferators-activated receptors (PPARs) and associated
transcription factors in colon cancer: Reduced expression of PPARγ-coactivator 1 (PGC1). Cancer Lett. 203, 25-33.
Firth, J. D., Ebert, B. L., Pugh, C. W., and Ratcliffe, P. J. (1994). Oxygenregulated control elements in the phosphoglycerate kinase 1 and lactate dehydrogenase A
genes: similarities with the erythropoietin 3’enhancer. PNAS. 91, 6496-6500.
Firth, J. D., Ebert, B. L., and Ratcliffe, P. J. (1995). Hypoxic regulation of
lactate dehydrogenase A: interaction between hypoxia-inducible factor 1 and cAMP
response elements. J Biol Chem. 270, 21021-21027.
Folch, J., Lees, M., and Sloane Stanley, G.H. (1957). A simple method for the
isolation and purification of total lipids from animal tissues. J Biol Chem. 226, 497-509.
Gillies, R.J., and Gatenby, R.A. (2007). Adaptive landscapes and emergent
phenotypes: why do cancers have high glycolysis? J Bioenerg Biomembr. 39, 251-257.
Goffart, S., and Wiesner, R.J. (2003). Regulation and co-ordination of nuclear
gene expression during mitochondrial biogenesis. Exp Physiol. 88. 33-40.
Gottlob, K., Majewski, N., Kennedy, S., Kandel, E., Robey, R.B., and Hay,
N. (2001). Inhibition of early apoptotic events by Akt/PKB is dependent on the first
committed step of glycolysis and mitochondrial hexokinase. Genes Dev. 15, 1406-1418.
Ghafar, M.A., Anastasiadis, A.G., Chen, M.-W., Burchardt, M., Olsson,
L.E., Hui Xie, Benson, M.C. and Buttyan, R. (2003). Acute hypoxia increases the
aggressive characteristics and survival properties of prostate cancer cells. Prostate. 54,
58-67.
Graeber, T.G., Osmanian, C., Jacks, T., Housman, D.E., Koch, C.J., Lowe,
S.W., and Giaccia, A.J. (1996). Hypoxia-mediated selection of cells with diminished
apoptotic potential in solid tumours. Nature. 379, 88-91.
Green, D.R., and Chipuk, J.E. (2006). P53 and metabolism: Inside the TIGAR.
Cell. 126, 30-32.
Greijer, A.E., and van der Wall, E. (2004). The role of hypoxia inducible factor
1 (HIF-1) in hypoxias induced apoptosis. JCP. 57, 1009-1014.
80
Guppy, M. (2002). The hypoxic core: a possible answer to the cancer paradox.
Biochem Biophys Res Commun. 299, 676–80.
Hardman, W., Gansler, T., Schaffel, S., and Henninger, R. (1995). OA-519
immunostaining portends poor prognosis in ovarian cancer. Mod Pathol. 8, 90A.
Hayward, S.W., Haughney, P.C., Lopes, E.S., Danielpour, D., and Cunha,
G.R. (1999) The rat prostatic epithelial cell line NRP-152 can differentiate in vivo in
response to its stromal environment. Prostate. 39, 205-212.
Heemers, H., Maes, B., Foufelle, F., Heyns, W., Verhoeven, G., and Swinnen,
J.V. (2001) Androgens stimulate lipogenic gene expression in prostate cancer cells by
activation of the sterol regulatory element-binding protein cleavage activating
protein/sterol regulatory element-binding protein pathway. Molec Endo. 15, 1817-1828.
Hennessey, B.T., Smith, D.L., Ram, P.T., Lu, Y., Mills, G.B. (2005).
Exploiting the PI3K/AKT pathway for cancer drug discovery. Nat Rev Drug Discov. 4,
988-1004.
Hochachka, P.W., Rupert, J.L., Goldenberg, L., Gleave, M., and Kozlowski,
P. (2002). Going malignant: the hypoxia-cancer connection in the prostate. BioEssays.
24, 749-757.
Hoeben, A., Landuyt, B., Highley, M.S., Wildiers, H., Van oosterom, A.T.,
and De Bruijn, E.A. (2004). Vascular endothelial growth factor and angiogenesis.
Pharmacol Rev. 56, 549-580.
Hollstein, M., Hergenhahn, M., Yang, Q., Bartsch, H., Wang, Z.-Q., and
Hainaut, P. (1999). New approaches to understanding p53 gene tumor mutation spectra.
Mutat. Res. 431, 199–209.
Hood, D.A. (2001). Contractile activity-induced mitochondrial biogenesis in
skeletal muscle. J Appl Physiol. 90, 1137-1157.
Horoszewicz, J.S., Leong, S.S., Chu, T.M., Wajsman, Z.L., Friedman, M.,
Papsidero, L., Kim, U., Chai, L.S., Kakati, S., Arya, S.K. and Sandberg, A.A.
(1980). The LNCaP cell line--a new model for studies on human prostatic carcinoma.
Prog Clin Biol Res. 37, 115-132.
Hua, X., Goldstein, J.L., Brown, M.S., and Hobbs, H.H. (1995). Structure of
81
the human gene encoding sterol regulatory element binding protein-1 (SREBPF1) and
localization of SREBPF1 and SREBPF2 to chromosomes 17p11.2 and 22q13. Genomics.
25, 667-673.
Huang, L.E., Arany, Z., Livingston, D.M., and Bunn, H.F. (1996). Activation
of hypoxia-inducible transcription factor depends primarily upon redox-sensitive
stabilization of its alpha subunit. J Biol Chem. 271, 32253-32259.
Isaacs, W.B., Carter, B.S., and Ewing, C.M. (1991). Wild-type p53 suppresses
growth of human prostate cancer cells containing mutant p53 alleles. Cancer Res. 51,
4716-4720.
Ivan, M., Kondo, K., Yang, H., Kim, W., Valiando, J., Ohh, M., Salic, A.,
Asara, J.M. Lane, W.S., and Kaelin, W.G. Jr. (2001). HIFalpha targeted for VHLmediated destruction by proline hydroxylation: implications for O2 sensing. Science.
292, 464-468.
Jiang, W.G., Douglas-Jones, A., and Mansei, R.E. (2003). Expression of
peroxisome-proliferator activated receptor-gamma (PPARγ) and the PPARγ co-activator,
PGC-1, in human breast cancer correlates with clinical outcomes. Int J Cancer. 106, 752757.
Kaighn, M.E., Narayan, K.S., Ohnuki, Y., Lechner, J.F., and Jones, L.W.
(1979). Establishment and characterization of a human prostatic carcinoma cell line (PC3). Invest Urol. 17, 16-23.
Kaighn, M.E., Lechner, J.F., Babcock, M.S., Marnell, M., Ohnuki, Y., and
Narayan, K.S. (1980). The Pasadena cell lines. Prog Clin Biol Res. 37, 85-109.
Kallinowski, F., Schaefer, C., Tyler, G., and Vaupel, P. (1989). In vivo targets
of recombinant human tumour necrosis factor-alpha: blood flow, oxygen consumption
and growth of isotransplanted rat tumours. Br J Cancer. 60, 555-560.
Kennedy, S.G., Kandel, E.S., Cross, T.K., and Hay, N. (1999). Akt/protein
kinase B inhibits cells death by preventing the release of cytochrome c from
mitochondria. MCB. 19, 5800-5810.
Kim, J. –Y., and Park, J. –H. (2003). ROS-dependent caspase-9 activation in
hypoxic cell death. FEBS Letters. 549, 94-98.
Kim, K., Yoshida, D., and Teramoto, A. (2005). Expression of hypoxiainducible factor 1α and vascular endothelial growth factor in pituitary adenomas. Endocr
82
Pathol. 16, 115-121.
Klingenspor, M., Dickopp, A., Heldmaier, G., and Klaus, S. (1996). Short
photoperiod reduces leptin gene expression in white and brown adipose tissue og
Djungarian hamster. FEBS Lett. 399. 290-294.
Kuhajda, F.P., Jenner, K., Wood, F.D., Hennigar, R.A., Jacobs, L.B., Dick,
J.D., and Pasternack, G.R. (1994). Fatty acid synthesis: a potential selective target for
antineoplastic therapy. PNAS. 91, 6379-6383.
Kuhajda, F. (2000). Fatty-acid synthase and human cancer: new perspectives on
its role in tumor biology. Nutrition.16, 202-208.
Leary, S.C., Kaufman, B.A., Pellecchia, G., Guercin, G.-H., Mattman, A.,
Jaksch, M., and Shobridge, E.A. (2004). Human SCO1 and SCO2 have independent,
cooperative function in copper delivery to cytochrome c oxidase. Human Mol Gen. 13,
1839-1848.
Lee, S.O., Lou, W., Hou, M., Onate, S.A., and Gao, A.C. (2003). Interleukin-4
enhances prostate-specific antigen expression by activation of the androgen receptor and
Akt pathway. Oncogene. 22, 7981–7988.
Lenka, N., Vijayasarathy, C., Mullick, J., and Avadhani, N.G. (1998).
Structural organization and transcription regulation of nuclear genes encoding the
mammalian cytochrome c oxidase complex. Prog Nucl Acid Res Mol Bio. 61, 309-344.
Lin, K., Sherrington, P.D., Dennis, M., Matrai, Z., Cawley, J.C., and Pettitt,
A.R. (2002). Relationship between p53 dysfunction, CD38 expression, and IgV(H)
mutation in chronic lymphocytic leukemia. Blood. 100, 1404-1409.
Masayesva, B.G., Mambo, E., Taylor, R.J., Goloubeva, O.G., Zhou, S.,
Cohen, Y., Minhas, K., Koch, W., Sciubba, J., Alberg, A.J., Sidransky, D., and
Califano, J. (2006). Mitochondrial DNA content increase in response to cigarette
smoking. Cancer Epidemiol Biomarkers Prev. 15, 19-24.
Matoba, S., Kang, J. -G., Patino, W.D., Wragg, A., Boehm, M., Gavrilova,
O., Hurley, P.J., Bunz, F., and Hwang, P.M. (2006). P53 regulated mitochondrial
respiration. Science. 312, 1650-1653.
Maxwell, P.H., Dachs, G.U., Gleadle, J.M., Nicholls, L.G., Harris, A.L.,
Stratford, I.J., Hankinson, O., Pugh, C.W., and Ratcliffe, P.J. (1997). Hypoxiainducible factor-1 modulates gene expression in solid tumors and influences both
83
angiogenesis and tumor growth. PNAS. 94, 8104–8109.
Medes, G., Thomas, A., and Weinhouse, S. (1953). Metabolism of neoplastic
tissue. IV. A study of lipid synthesis in neoplastic tissue slices in vitro. Cancer Res. 13,
27-29.
Mickey, D.D., Stone, K.R., Wunderli, H., Mickey, G.H., and Paulson, D.F.
(1980). Characterization of a human prostate adenocarcinoma cell line (DU 145) as a
monolayer culture and as a solid tumor in athymic mice. Prog Clin Biol Res. 37, 67-84.
Miralpeix, M., Azcon-Bieto, J., Bartrons, R., and Argiles, J.M. (1990). The
impairment of respiration by glycolysis in the Lewis lung carcinoma. Cancer Lett. 50,
173-178.
Miserez, A.R., Cao, G., Probst, L.C., and Hobbs, H.H. (1997). Structure of the
human gene encoding sterol regulatory element binding protein 2 (SREBPF2).
Genomics. 40, 31-40.
Mohler, J.L., Gregory, C.W., Ford, O.H. 3rd, Kim, D., Weaver, C.M.,
Petrusz, P., Wilson, E.M., and French, F.S. (2004). The androgen axis in recurrent
prostate cancer. Clin Cancer Res. 10, 440–448.
Moreau, K., Dizin, E., Ray, H., Luquain, C., Lefai, E., Foufelle, F., Billaud,
M., Lenoir, G.M., and Venezia, N.D. (2006). BRCA1 Affects lipid synthesis through
its interaction with acetyl-CoA carboxylase. J Biol Chem. 281, 3172-3181.
Moreno-Sanchez, R., Rodriguez-Enriquez, S., Marin-Hernandez, A., and
Saavedra, E. (2007). Energy metabolism in tumour cells. FEBS Lett. 274, 1393-1418.
Moscow, J.A., and Cowen, K.H. (1988). Multidrug resistance. J Natl Cancer
Inst. 80, 14-20.
Moyes, C.D., Mathieu-Costello, O.A., Tsuchiya, N., Filburn ,C., and
Hansford, R.G. (1997). Mitochondrial biogenesis during cellular proliferation. Am J
Physiol Cell Physiol. 272, C1345-C1351.
Neufeld, G., Cohen, T., Gengrinovitch, S., and Poltorak, Z. (1999). Vascular
endothelial growth factor (VEGF) and its receptors. FASEB J. 13, 9-22.
Ookhtens, M., Kanna, R., Lyon, I., and Baker, N. (1984). Liver and adipose
tissue contributions to newly formed fatty acids in an ascites tumour. Am J Physiol
Regul Integr Comp Physiol. 247, 146-153.
84
Park, S.Y., Kim, Y.J., Gao, A.C., Onate, S.A., Hidalgo, A.A., Ip, C., Park,
E.M., Yoon, S.Y., and Park, S.M. (2006). Hypoxia increases androgen receptor activity
in prostate cancer cells. Cancer Res. 66, 5121-5129.
Pelicano, H., Xu, R.-H., Du, M., Feng, L., Sasaki, R., Carew, J.S., Hu, Y.,
Ramdas, L., Hu, L., Keating, M.J., Zhang, W., Plunkett, W., and Huang, P. (2006).
Mitochondrial respiration defects in cancer cells cause activation of Akt survival
pathway through a redox-mediated mechanism. J Cell Biol. 175, 913-923.
Petrylak, D.P., Tangen, C.M., Hussain, M.H., Lara, P.N., Jones, J.A., Taplin,
M.E., Burch, P.A., Berry, D., Moinpour, C., Kohli, M., Benson, M.C., Small, E.J.,
Raghavan, D., and Crawford, E.D. (2004). Docetaxel and estramustine compared with
mitoxantrone and prednisone for advanced refractory prostate cancer. N Eng J Med. 351,
1513-1520.
Pizer, E.S., Lax, S.F., Kuhajda, F.P., Pasternack, G.R., and Kurman, R.J.
(1998). Fatty acid synthase expression in endometrial carcinoma: correlation with cell
proliferation and hormone receptors. Cancer. 83, 528-537.
Puigserver, P., Wu, Z., Park, C.W., Graves, R., Wright, M., and Spiegelman,
B.M. (1998). A cold-inducible coactivator of nuclear recptors linked to adaptive
thermogenesis. Cell. 92, 829-839.
Rashid, A., Pizer, E.S., Moga, M., Milgraum, L.Z., Zahurak, M., Pasternack,
G.R., Kuhajda, and F.P., Hamilton, S.R. (1997). Elevated expression of fatty acid
synthase and fatty acid synthetic activity in colorectal neoplasia. Amer J Pathol. 150,
201-208.
Rice, G.C., Hoy, C., and Schimke, R.T. (1986). Transient hypoxia enhances the
frequency of dihydrofolate reductase gene amplification in Chinese hamster ovary cells.
PNAS. 83, 5978-5982.
Robey, R.B., and Hay, N. (2005). Mitochondrial hexokinases: guardians of the
mitochondria. Cell Cycle. 4, 654-658.
Rodriguez-Enriquez, S., Kim, I., Currin, R.T., and Lemasters, J.J. (2006).
Tracker dyes to probe mitochondrial autophagy (mitophagy) in rat hepatocytes.
Autophagy. 2, 39-46.
Rossi, S., Graner, E., Febbo, P., Weinstein, L., Bhattacharya, N., Onody, T.,
Bubley, G., Balk, S., and Loda, M. (2003). Fatty acid synthase expression defines
85
distinct molecular signatures in prostate cancer. Mol Cancer Res. 1, 707-715.
Robey, R.B., and Hay, N. (2005). Mitochondrial hexokinases, novel mediators
of the antiapoptotic effects of growth factors and Akt. Oncogene. 25, 4683-4696.
Saikumar, P., Dong, Z., Patel, Y., Hall, K., Hopfer, U., Weinberg, J.M.,
Venkatachalam, M.A. (1998). Role of hypoxia-induced Bax translocation and
cytochrome c release in reoxygenation injury. Oncogene. 17, 3401-3415.
Scarpulla, R.C. (1997). Nuclear control of respiratory chain expression in
mammalian cells. J Bioenerg Biomembr. 29, 109-119.
Seagroves, T.N., Ryan, H.E., Lu, H., Wouters, B.G., Knapp, M., Thibault, P.,
Laderoute, K., and Johnson, R.S. (2001). Transcription factor HIF-1 is a necessary
mediator of the Pasteur effect in mammalian cells. Mol Cell Biol. 21, 3436-3444.
Semenza, G.L., and Wang, G.L. (1992). 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. 12, 5447–5454.
Semenza, G.L. (1994). Regulation of erythropoietin production. New insights
into molecular mechanisms of oxygen homeostasis. Hematol Oncol Clin North Am. 8,
863–884.
Semenza, G. L., Roth, P. H., Fang, H.-M., and Wang, G. L. (1994).
Transcriptional regulation of genes encoding glycolytic enzymes by hypoxia-inducible
factor 1. J Biol Chem. 269, 23757-23763.
Semenza, G.L., Jiang, B.-H., Leung ,S.W., Passantino, R., Concordet, J.-P.,
Maire, P., and Giallongo, A. (1996). Hypoxia response elements in the aldolase a,
enolase 1, and lactate dehydrogenase a gene promoters contain essential binding sites for
hypoxia-inducible factor 1. J Biol Chem. 271, 32529-32537.
Semenza, G.L. (2000). Hypoxia, clonal selection, and the role of HIF-1 in
tumour progression. Crit Rev Biochem Mol Biol. 35, 71-103.
Semenza, G.L. (2007). Hypoxia-inducible factor 1 pathway. Science Signaling.
2007, cm8.
Shamash, J., Dancey, G., Barlow, C., Wilson, P., Ansell, W., and Oliver, R.T.
(2005). Chlorambucil and lomustine (CL56) in absolute hormone refractory prostate
cancer: re-introduction of endocrine sensitivity an unexpected finding. Br J Cancer. 92.
86
36-40.
Shaw, G.L., Wilson, P., Cuzick, J., Prowse, D.M., Goldenberg, S.L., Spry,
N.A., and Oliver, T. (2007). International study into the use of intermittent hormone
therapy in the treatment of carcinoma of the prostate: a meta-analysis of 1446 patients.
BJU. 99, 1056-1065.
Shurbaji, M.S., Kuhajda, F.P., Pasternack, G.R., and Thurmond, T.S.
(1992). Expression of oncogenic antigen 519 (OA-519) in prostate cancer is a potential
prognostic indicator. Amer J Clin Pathol. 97, 686-691.
Smalley, K.S.M., Brafford, P.A., and Heryln, M. (2005). Selective
evolutionary pressure from the tissue microenvironment drives tumour progression.
Semin Cancer Biol. 15, 451-459.
Stambolic, V., Suzuki, A., de la Pompa, J.L., Brothers, J.M., Mirtsos, C.,
Sasaki, T., Ruland, J., Penninger, J.M., Siderovski, D.P., and Mak, T.W. (1998).
Negative regulation of PKB/Akt—dependent cell survival by the tumor suppressor
PTEN. Cell. 95, 29-39.
Stump, C.S., Short, K.R., Bigelow, M.L., Schimke, J.M., and Nair, S. (2003).
Effect of insulin on human skeletal muscle mitochondrial ATP production, protein
synthesis, and mRNA transcripts. PNAS. 100: 7996-8001.
Sutherland, R.M. (1988). Cell and environment interactions in tumour
microregions: the multicell spheroid model. Science. 240, 177-184.
Swinnen, J.V., Veldhoven, P.P., Esquenet, M., Heyns, W., and Verhoeven, G.
(1996). Androgens markedly stimulate the accumulation of neutral lipids in the human
prostatic adenocarcinoma cell line LNCaP. Endocrinology. 137, 4468-4474.
Swinnen, J.V., Ulrix, W., Heyns, W., and Verhoeven, G. (1997). Coordinate
regulation of lipogenic gene expression by androgens: Evidence for a cascade
mechanism involving sterol regulatory element binding proteins. 94, 12975-12980.
Swinnen, J.V., Heemerss, H., Deboel, L., Foufelle, F., Heyns, W., and
Verhoeven, G. (2000a). Stimulation of tumour-associated fatty acid synthase expression
by growth factor activation of the sterol regulatory element-binding protein pathway.
Oncogene. 19, 5173-5181.
Swinnen, J.V., Vanderhoydonc, F., Elgamal, A.A., Eelen, M., Vercaeren, I.,
Joniau, S., Van Poppel, H., Baert, L.,Goosens, K., Heyns, W., and Verhoeven, G.
(2000b). Selective activation of the fatty acid synthesis pathway in human prostate
87
cancer. Int J Cancer. 88, 176-179.
Swinnen, J.V., Roskams, T., Joniau, S., Van Poppel, H., Oyen, R., Baert, L.,
Heyns, W., and Verhoeven, G. (2002). Overexpression of fatty acid synthase is an early
and common event in the development of prostate cancer. Int J Cancer. 98, 19-22.
Swinnen, J.V., Van Veldhoven, P.P., Timmermans, L., De Schrijver, E.,
Brusselmans, K., Vanderhoydonc, F., Van de Sande, T., Heemerss, H., Heyns, W.,
and Verhoeven, G. (2003). Fatty acid synthase drives the synthesis of phospholipids
partitioning into detergent-resistant membrane microdomains. Biochem Biophys Res
Commun. 302, 898-903.
Swinnen, J., Brusselmans, K., and Verhoeven, G. (2006). Increased
lipogenesis in cancer cells: new players, novel targets. Curr Opin Clin Nutr Metab Care.
9, 358-365.
Taplin, M.E., Bubley,G.J., Shuster, T.D., Frantz, M.E., Spooner, A.E.,
Ogata, G.K., Keer, H.N., and Balk, S.P. (1995). Mutation of the androgen-receptor
gene in metastatic androgen-independent prostate cancer. N Engl J Med. 332,1393–1398.
Taplin, M.E., Rajeshkumar, B., Halabi, S., Werner, C.P., Woda, B.A., Picus,
J., Stadler, W., Hayes, D.F., Kantoff, P.W., Vogelzang, N.J. and Small, E.J. (2003).
Androgen receptor mutations in androgen-independent prostate cancer: Cancer and
Leukemia Group B Study 9663. J Clin Oncol. 21, 2673–26788.
Takahashi, Y., Bucana, C.D., Cleary, K.R., and Ellis, L.M. (1998). P53,
vessel count, and vascular endothelial growth factor expression in human colon cancer.
Int J Cancer. 79, 34-38.
Vaupel, P., Kelleher, D.K., and Hockel, M. (2001). Oxygen status of malignant
tumours: pathogenesis of hypoxia and significance for tumour therapy. Semin Oncol 2.
Suppl 8, 29-35.
Veldscholte, J., Berrevoets, C.A., Brinkmann, A.O., Grootegoed, J.A., and
Mulder, E. (1992). Anti-androgens and the mutated androgen receptor of LNCaP cells:
differential effects on binding affinity, heat-shock protein interaction, and transcription
activation. Biochemistry. 31, 2393-2399.
Visakorpi, T., Hyytinen, E., Koivisto, P., Tanner, M., Keinanen, R.,
Palmberg, C., Palotie, A., Tammela, T., Isola, J., and Kallioniemi, O.P. (1995). In
vivo amplification of the androgen receptor gene and progression of human prostate
cancer. Nat Genet. 9, 401–406.
88
Vivanco, I., and Sawyers, C.L. (2002). The phosphatidylinositol 3-Kinase AKT
pathway in human cancer. Nature Rev Cancer. 2, 489-501.
Vogelstein, B., Lane, D. and Levine, A.J. (2000). Surfing the p53 network.
Nature. 408, 307–310.
Vogelstein, B., and Kinzer, W. (2004). Cancer genes and the pathways they
control. Nat Med. 10, 789-799.
Wallace, D.C. (1999). Mitochondrial diseases in man and mouse. Science. 283,
1482–1488.
Warburg, O., Wind, F., Negelein, E., and den Stoffwechsel, U. (1926). Von
Tumouren im Korper. Klin Woch. 5, 829-832.
Warburg, O. (1930). The Metabolism of Tumours. Constable Press, London.
Wei, M.C., Zong, W.- X., Cheng, E. H.- Y., Lindsten, T., Panoutsakopoulou,
V., Ross, A.J., Roth, K.A., MacGregor, G.R., Thompson, C.B., and Korsmeyer, S.J.
(2001). Proapoptotic BAX and BAK: A requisite gateway to mitochondrial dysfunction
and death. 292, 727-730.
Weinhouse, S. (1956). On respiratory impairment in cancer cells. Science. 124,
267-269.
Welsh, J.B., Sapinoso, L.M., Su, A.I., Kern, S.G., Wang-Rodriquez, J.,
Moskaluk, C.A., Frierson, H.F., and Hampton, G.M. (2001). Analysis of gene
expression identifies candidate markers and pharmacological targets in prostate cancer.
Cancer Res. 61, 5974-5978.
Wilding, G., Zugmeier, G., Knabbe, C., Flanders, K., and Gelmann, E.
(1989). Differential effects of transforming growth factor-ß on human prostate cancer
cells in vitro. Mol Cell Endocrinol. 62, 79-87.
Wu, Z., Puigserver, P., Andersson, U., Zhang, C., Adelmant, G., Mootha, V.,
Troy, A., Cinti, S., Lowell, B., Scarpulla, R.C., and Spiegelman, B.M. (1999).
Mechanisms controlling mitochondrial biogenesis and respiration through the
thermogenic coactivator PGC-1. Cell. 98, 115-124.
Xu, R. H., Pelicano, Y., Zhou, J.S., Carew, L., Feng, K.N. and Bhalla, M.J.
Keating, and P. Huang. (2005). Inhibition of glycolysis in cancer cells: a novel strategy
to overcome drug resistance associated with mitochondrial respiratory defect and
89
hypoxia. Cancer Res. 65, 613–621.
Yoon, J.C., Puigserver, P., Chen, G., Donovan, J., Wu, Z., Rhee, J.,
Adelmant, G., Stafford, J., Kahn, C.R., Granner, D.K., Newgard, C.B., Spiegelman,
B.M. (2001). Control of hepatic gluconeogenesis through the transcriptional coactivator
PGC-1. Nature. 413, 131-138.
Young, C.D., and Anderson, S.M. (2008). Sugar and fat- that’s where it’s at:
metabolic changes in tumors. Br Cancer Res. 10, 202-211.
Zhang, G.Y., Ahmed, N., Riley, C., Oliva, K., Barker, G., Quinn, M.A., and
Rice, G.E. (2005). Enhanced expression of peroxisome proliferator-activated receptor
gamma in epithelial ovarian carcinoma. BJC. 92, 113-119.
Zhong, H., Agani, F., Baccala, A.A., Laughner, E., Rioseco-Camacho, N.,
Isaacs, W.B., Simons, J.W., and Semenza, G.L. (1998). Increased expression of
hypoxia inducible factor-1a in rat and human prostate cancer. Cancer Res. 58, 52805284.
Zu, X., and Guppy, M. (2004). Cancer metabolism: facts, fantasy and fiction.
Biochem Biophys Res Commun. 313, 459-465.
90
Appendix 1
Supplementary Figure 1. Integration of Cellular Pathways Involved in Aerobic
Glycolysis in Cancer Cells.
Some cancer cells depend primarily on glycolysis for cellular ATP production. This may
be due to 1) selection under hypoxic conditions, 2) mitochondrial dysfunction, 3)
activation of Akt, or 4) inactivation of p53. Under hypoxic conditions, HIF-1 stimulates
the transcription of its targets involved in glucose uptake and glycolysis. Similarly,
oxidative impairment might lead to a cell with an increased glycolytic rate. Activation of
Akt stimulate glycolysis without affecting cellular respiration which may also explain
aerobic glycolysis. Lastly, inactivation of p53 may be responsible for aerobic glycolysis
via the proteins TIGAR and SCO2. TIGAR is involved in shunting glucose-6-phosphate
to the pentose phosphate pathway. SCO2 is involved in complex IV holoenzyme
formation. These proteins are expressed in a p53 dependent manner.
Figure obtained from Young and Anderson (2008).
91
Supplementary Table 1. Raw cycle thresholds of real-Time PCR data
comparing LNCaP, DU145, and PC3 cells (Figure 7)
SREBP1a
SREBP-2
SREBP1c
SCAP
FAS
Acc1
HKII
HIF1a
VEGF65
VEGF212
ALDOa
PKm
LDHa
PGC-1a
NRF-2
COXIV
CS
COXI
LNCaP
mean
sem
DU145
mean
sem
PC3
mean
sem
1.87
5.96
0.16
1.03
2.66
19.42
0.17
1.35
3.78
13.71
0.25
0.65
12.26
4.82
24.23
3.00
1.14
6.17
1.19
1.04
323.33
1.40
178.88
0.22
1.78
87.43
60.52
888.16
0.95
2.36
0.36
13.22
3.92
20.08
0.28
3.41
0.10
1.20
1.17
137.57
0.12
54.54
0.09
18.88
61.40
547.94
0.42
2.74
34.55
197.68
0.03
0.37
0.07
4.39
17.60
252.93
14.85
162.73
120.64
2429.45
92
0.29
1.67
0.07
1.99
18.22
1.71
2.51
20.61
1.91
0.14
3.53
0.18
0.26
8.29
0.14
24.82
113.91
8.04
2.29
10.37
0.42
1.06
3.09
0.23
99.61
1100.59
431.44
0.57
18.65
6.97
38.80
754.65
251.21
0.24
1.36
0.49
0.89
8.30
0.42
42.22
70.03
10.42
31.43
63.20
1.98
93.35
1558.66
92.26
Supplementary Table 2. Raw cycle thresholds of real-Time PCR data
comparing LNCaP and derived lines (Figure 11)
SREBP1a
SREBP-2
SREBP1c
SCAP
FAS
Acc1
HKII
HIF1a
VEGF65
VEGF212
ALDOa
PKm
LDHa
PGC-1a
NRF-2
COXIV
CS
COXI
LNCaP
mean
sem
HPLNCaP
mean
1.87
5.96
0.16
1.03
2.97
8.95
12.26
4.82
24.23
3.00
1.14
6.17
1.19
1.04
323.33
1.40
178.88
0.22
1.78
87.43
60.52
888.16
sem
LNCaPluc
mean
sem
0.34
1.38
1.83
5.71
0.51
1.16
0.95
11.86
1.51
0.21
0.36
9.11
0.68
12.26
3.92
16.24
0.59
4.74
0.28
4.27
0.47
1.71
0.10
1.64
0.25
2.07
1.17
16.59
0.69
118.92
0.12
3.49
0.68
2.42
0.09
4.52
1.01
1.40
61.40
612.38
82.28
871.26
0.42
2.68
0.26
18.38
34.55
310.29
37.93
584.80
0.03
0.53
0.07
0.12
0.07
3.22
0.24
24.26
17.60
112.28
4.66
153.80
14.85
69.10
1.63
71.72
120.64
4350.15
1010.16
1400.55
93
0.06
0.67
0.62
0.15
0.16
9.00
0.14
0.12
57.57
1.60
44.85
0.02
1.95
12.35
3.29
57.59
Supplementary Table 3. Raw cycle thresholds of real-Time PCR data
comparing HP-LNCaPs and xenografts (Figure 12)
HPHP
LNCaP
xenograft
mean
sem
mean
sem
SREBP1a
2.97
0.34
2.16
0.21
SREBP-2
8.95
1.38
16.32
2.66
SREBP11.86
1.51
5.87
0.51
1c
SCAP
9.11
0.68
44.13
6.75
FAS
16.24
0.59
9.80
2.13
Acc1
4.27
0.47
2.32
0.29
HKII
1.64
0.25
2.34
0.86
HIF1a
16.59
0.69
28.89
4.09
VEGF65
3.49
0.68
10.59
2.90
VEGF212
4.52
1.01
9.09
2.16
ALDOa
612.38
82.28
150.02
32.40
PKm
2.68
0.26
1.41
0.57
LDHa
310.29
37.93
127.78
64.81
PGC-1a
0.53
0.07
3.85
1.89
NRF-2
3.22
0.24
4.94
1.06
COXIV
112.28
4.66
353.53
120.23
CS
69.10
1.63
49.35
11.33
COXI
4350.15
1010.16
17001.49
5153.78
94
Supplementary Table 4. Raw cycle thresholds of real-Time PCR data
comparing LNCaP-luc cells and xenografts (Figure 13)
LNCaP-luc
kidney
LNCaP-luc
LNCaP-luc leg
mean
sem
mean
sem
mean
sem
SREBP1a
1.83
0.51
1.30
0.10
1.73
0.23
SREBP-2
5.71
1.16
5.69
0.32
9.14
0.01
SREBP1c
0.21
0.06
0.11
0.02
0.17
0.02
SCAP
12.26
0.67
14.15
1.00
13.63
1.17
FAS
4.74
0.62
4.29
0.34
7.78
0.78
Acc1
1.71
0.15
1.24
0.15
1.22
0.14
HKII
2.07
0.16
3.88
0.45
3.78
0.65
HIF1a
118.92
9.00
63.85
9.21
50.90
3.02
VEGF65
2.42
0.14
2.71
0.22
4.58
0.66
VEGF212
1.40
0.12
1.24
0.14
2.02
0.26
ALDOa
871.26
57.57
775.26
37.67
1131.93
263.94
PKm
18.38
1.60
28.90
2.05
33.90
3.23
LDHa
584.80
44.85
395.42
27.78
355.50
21.23
PGC-1a
0.12
0.02
0.25
0.05
0.06
0.01
NRF-2
24.26
1.95
18.81
1.79
16.47
0.83
COXIV
153.80
12.35
158.32
10.29
97.70
5.65
CS
71.72
3.29
40.03
2.17
45.11
2.37
COXI
1400.55
57.59
3023.87
436.86
6261.51
1530.01
95
Supplementary Table 5. Raw cycle thresholds of real-Time PCR data
comparing LNCaP, DU145, and PC3 cells following low oxygen treatment (Figure
14)
LNCaP
control
hypoxia
anoxia
azide
DU145
control
hypoxia
anoxia
azide
PC3
control
hypoxia
anoxia
azide
SREBP
VEGF
1a
mean
1.12
1.74
sem
0.11
0.07
mean
1.36
1.38
sem
0.09
0.09
mean
3.62
1.83
sem
0.35
0.08
mean
2.01
0.58
sem
0.25
0.03
SREBP
VEGF
1a
mean
8.14
1.14
sem
0.57
0.05
mean
24.56
0.99
sem
2.31
0.10
mean
21.26
0.94
sem
5.01
0.08
mean
14.06
0.34
sem
5.22
0.04
SREBP
VEGF
1a
mean
3.29
1.05
sem
0.40
0.08
mean
14.47
3.01
sem
0.89
0.38
mean
19.43
2.19
sem
3.02
0.08
mean
5.28
0.35
sem
0.66
0.03
SREBP1c
SCAP
FAS
ACC1
HKII
2.22
6.05
12.70
3.78
0.92
0.19
0.36
1.82
0.15
0.05
1.26
4.82
6.36
2.28
1.45
0.09
0.13
0.42
0.10
0.11
1.34
5.52
5.77
2.02
3.45
0.02
0.46
0.87
0.11
0.28
0.53
6.26
4.79
2.21
0.93
0.09
1.13
0.91
0.44
0.26
SREBP1c
SCAP
FAS
ACC1
HKII
1.81
5.86
10.57
1.78
1.55
0.42
0.47
1.44
0.12
0.16
2.75
5.26
9.90
1.58
2.00
0.32
0.33
1.38
0.12
0.14
2.12
4.38
7.72
1.21
2.80
0.41
0.37
0.92
0.11
0.27
1.55
3.90
3.96
1.31
1.66
0.93
0.08
0.30
0.08
0.12
SREBP1c
SCAP
FAS
ACC1
HKII
0.11
6.91
2.24
0.79
1.55
0.01
0.47
0.27
0.04
0.11
0.29
7.29
3.04
0.80
10.83
0.06
0.85
0.15
0.05
1.04
0.35
6.22
2.69
0.68
12.47
0.06
1.05
0.27
0.08
1.58
0.08
6.63
1.17
0.69
3.15
0.02
0.41
0.23
0.03
0.59
96