1. health and fitness
2. disease
3. cancer

REGULATION OF THE GABPβ PROMOTER AND
THE ROLE OF NRF-1 IN MAMMARY EPITHELIAL
MORPHOGENESIS
by
Christina L. Lamparter
A thesis submitted to the Graduate Program in Biochemistry within
the Department of Biomedical and Molecular Sciences
in conformity with the requirements for
the degree of Master of Science
Queen’s University
Kingston, Ontario, Canada
September, 2012
Copyright © Christina L. Lamparter, 2012
Abstract
Decreased expression of the tumor suppressor gene BRCA1 (breast cancer 1, early onset) is
frequently observed in sporadic breast tumors, with the decrease not attributed to mutations or
hypermethylation of the promoter. A critical regulator of the BRCA1 promoter is the GA-Binding
Protein (GABP), a heterotetramer of GABPα and GABPβ. Previous analysis of the GABPβ
promoter revealed a regulatory multi-protein complex containing Nuclear Respiratory Factor 1
(NRF-1), which was aberrant in SK-BR-3 cells, resulting in decreased GABPβ and BRCA1
expression. To identify the unknown co-regulators of the NRF-1-containing complex at the
GABPβ promoter, an immobilized-template assay containing the complex binding site was
utilized. Despite optimization of binding and elution conditions through variations in salt content,
Mg2+ concentration and pH, non-specific DNA-binding proteins were present in the column
eluate. Further experimentation is therefore required to distinguish between non-specific proteins
and complex co-activators. As this complex containing NRF-1 is also able to modulate BRCA1, a
regulator of luminal progenitor differentiation, a defect in NRF-1 or complex co-activators could
contribute to the abnormal morphogenesis and differentiation of BRCA1-deficient tumors. We
therefore investigated the role of NRF-1 in mammary cell morphogenesis and its link to the
GABP/BRCA1 pathway. As both GABPβ and NRF-1 are known regulators of nuclear-encoded
mitochondrial proteins, this association also provides a link between mitochondrial metabolism
and breast differentiation, both of which are frequently disrupted in breast cancer. An inducible
lentiviral system was used to generate NRF-1 knockdown cell lines to examine its effect on
morphogenesis. Growth in 3D culture resulted in abnormal cell structures having impaired cell
polarization and lumen formation. In monolayer culture, prolonged NRF-1 knockdown did not
result in decreased BRCA1 or GABP expression. However, these cells did display notable
mitochondrial dysfunction, accompanied by the downregulation of several NRF-1 target genes
involved in mitochondrial biogenesis including Tfam and cytochrome c. These results suggest a
ii
role for NRF-1 in mediating mammary morphogenesis through maintenance of functional
mitochondria. Further investigation into the role of the NRF-1-containing complex at the GABPβ
promoter during differentiation might also provide insight into the altered cell metabolism and
differentiation observed in cancer cells.
iii
Acknowledgements
First and foremost, I would like to thank Dr. Christopher Mueller for allowing me to
pursue my MSc. under his guidance and supervision. He has supported me throughout this thesis
with his patience and knowledge while also allowing me to work in my own way.
In the completion of this degree I have been blessed with a friendly and cheerful group of
Mueller Lab members, both past and present. Thank you all for providing support and friendship,
and helping me to develop a healthy fitness balance; benefiting both body and mind. I would
especially like to thank Rachael Klinoski for patiently teaching me the basics, and I attribute the
level of my Masters degree to her encouragement and assistance. I would also like to thank
Rachael Klinoski, Alyssa Cull, Heather Ritter, Crista Thompson and Kirsten Nesset for their
input, moral support and friendship.
Thanks to my supervisory committee members Dr. Scott Davey and Dr. David LeBrun
for their insight and helpful suggestions, and thank you to the Canadian Breast Cancer
Foundation – Ontario Region for their financial support of my project.
Most importantly, none of this would have been possible without the love and patience of
my family. My family, to whom this dissertation is dedicated, has been a constant source of love,
concern, strength and support throughout all my studies at University. I would like to express my
heart-felt gratitude to my family, thank you for the emotional support, encouragement, and faith
in my abilities.
iv
Table of Contents
Abstract ............................................................................................................................................ ii
Acknowledgements ....................................................................................................................... ..iv
List of Figures……………………………………………………………………………………viii
List of Abbreviations………………………………………………………………………………x
CHAPTER 1 Introduction and Literature Review ............................................................... ………1
1.1 Breast Cancer………………………………………………………………………….1
1.2 The Human Mammary Gland…………………………………………………………2
1.2.1 Structure and Development………………………………………………...2
1.2.2 Epithelial Hierarchy and Tumourigenesis……………………………….....3
1.2.3 The Cancer Stem Cell Theory……………………………………………...4
1.2.4 Luminal Progenitors as the Cell-of-Origin in Basal-Like Tumors…………5
1.3 BRCA1 and Tumourigenesis……………………………………………………….....6
1.3.1 Familial and Sporadic Breast Cancers……………………………………...6
1.3.2 BRCA1 as a Tumor Suppressor…………………………………………….7
1.3.3 BRCA1 in Development and Differentiation………………………………7
1.4 Regulation of the BRCA1 Gene……………………………………………………..10
1.5 GABP Regulation and Cellular Functions…………………………………………...12
1.6 Nuclear Regulators of Mitochondrial Metabolism…………………………………..14
1.6.1 GABP and NRF-1 as Regulators of Mitochondrial Biogenesis…………..15
1.6.2 NRF-1 during Development………………………………………………17
1.7 Mitochondrial Metabolism in Breast Cancer………………………………………...18
1.8 Rationale and Hypothesis……………………………………………………………19
CHAPTER 2 Isolation of a complex containing NRF-1 that regulates the GABPβ promoter…...20
2.1 Abstract………………………………………………………………………………20
2.2 Background…………………………………………………………………………..21
2.3 Methods………………………………………………………………………………23
2.3.1 Cells and cell culture………………………………………………………23
2.3.2 Design of dual biotin-labeled probes……………………….……………..23
2.3.3 Immobilized-template assays………………………………………….…..23
2.3.4 Immobilized-template competition assay…………………………………24
2.3.5 Western blot analysis……………………………………………………...25
2.3.6 Electrophoretic mobility shift assay (EMSA)………………………….….25
v
2.4 Results………………………………………………………………………………..25
2.4.1 Isolation of the NRF-1 containing protein complex that binds the GABPβ
promoter using an immobilized-template assay………………………….25
2.4.2 Non-specific binding proteins elute with NRF-1 from the immobilizedtemplate…………………………………………………………………...29
2.4.3 Optimization of the immobilized-template binding and elution conditions
……………………………………………………………………………29
2.4.4 Isolated fractions containing NRF-1 are not consistent with complex
formation………………………………………………………………….33
2.4.5 Evaluation of binding specificity to -268/-251 by fractions isolated using
the immobilized-template assay…………………………………………..35
2.5 Discussion……………………………………………………………………………38
2.6 Conclusions…………………………………………………………………………..43
2.7 Acknowledgements…………………………………………………………………..43
CHAPTER 3 The role of NRF-1 in BRCA1-mediated differentiation and mammary gland
tumourigenesis…………………………………………………………………………………....44
3.1 Abstract………………………………………………………………………………44
3.2 Background…………………………………………………………………………..45
3.3 Methods………………………………………………………………………………47
3.3.1 Cells and cell culture………………………………………………………47
3.3.2 Lentiviral production and transduction……………………………………47
3.3.3 Acini……………………………………………………………………….48
3.3.4 Preparation of whole cell lysates………………………………………….49
3.3.5 Western blot analysis……………………………………………………...49
3.3.6 Quantitative real-time PCR (qRT-PCR)…………………………………..49
3.3.7 Dual luciferase assay……………………………………………...………50
3.3.8 Mitochondrial function assay……………………………………………...51
3.4 Results………………………………………………………………………………..51
3.4.1 Mammary epithelial cells with decreased NRF-1 expression demonstrate
abnormal acinar morphogenesis………………………………………….51
3.4.2 NRF-1 depletion does not result in a long-term depression of
GABPβ/BRCA1 expression in monolayer cells………………………….59
3.4.3 Loss of NRF-1 is associated with mitochondrial dysfunction in mammary
epithelial cells…………………………………………………………….66
vi
3.4.4 The expression of genes involved in mitochondrial metabolism and
apoptosis are down-regulated with NRF-1 knockdown………………….68
3.5 Discussion……………………………………………………………………………70
3.6 Conclusions…………………………………………………………………………..74
3.7 Acknowledgements…………………………………………………………………..75
CHAPTER 4 General Discussion………………………………………………………………...76
4.1 Metabolic Adaptations in Tumors…………………………………………………...76
4.2 Metabolism and Differentiation of Embryonic Stem Cells………………………….76
4.3 Metabolic Reprogramming in Tumourigenesis……………………………..……….78
CHAPTER 5 Summary and Conclusions………………………………………………………...80
Reference List…………………………………………………………………………………….81
vii
List of Figures
Figure 1.1. Domain structures of GABPα, GABPβ1 isoforms and NRF-1……………………...13
Figure 1.2. NRF-1 and GABP coordinately regulate the expression of nuclear-encoded
mitochondrial proteins……………………………………………………………….16
Figure 2-1: Dual biotin-tagged probe for DNA affinity assay…………………………….……...27
Figure 2.2. Proteins bind non-specifically to the GABPβ promoter in an immobilized-template
assay………………………………….………………………………………………28
Figure 2.3. Non-specific proteins elute with NRF-1 in a gradient salt elution………….………..30
Figure 2.4. Optimization of elution conditions with Mg2+ and pH variation…………………….32
Figure 2.5. Fractions containing NRF-1 are not consistent with complex formation……………34
Figure 2.6. NRF-1 supplemented fractions do not mediate additional complex formation……...36
Figure 2.7. Complex binding is diminished upon binding site mutation……………….………...37
Figure 3.1. Inducible lentiviral mediated knockdown of NRF-1 protein.………………………...53
Figure 3.2. Inducible lentiviral mediated knockdown of NRF-1 mRNA………………………...54
Figure 3.3. NRF-1 knockdown impairs mammary cell differentiation in 184hTERTs………......56
Figure 3.4. NRF-1 knockdown impairs mammary cell differentiation in MCF-10As………...…57
Figure 3.5. NRF-1 deficient cells exhibit impaired lumen formation……...……………………..58
Figure 3.6. NRF-1 depletion does not result in a prolonged decrease of GABPα, GABPβ and
BRCA1 expression in monolayer 184hTERT cells……………….....………………60
Figure 3.7. NRF-1 depletion does not result in a prolonged decrease of GABPα, GABPβ and
BRCA1 expression in monolayer MCF-10A cells………...……………..………….61
Figure 3.8. Endogenous mRNA expression of GABPβ is reduced in MCF-10A cells...…………63
Figure 3.9. GABPβ promoter activity differs between 184hTERT and MCF-10A cells upon
NRF-1 depletion…………...…………………………………………………………65
Figure 3.10. Mitochondrial dysfunction results from NRF-1 knockdown………………….……67
viii
Figure 3.11. Decreased Tfam and cytochrome c expression is associated with NRF-1 knockdown
..……………………………………………………………………………………..69
ix
List of Abbreviations
53BP1
p53 Binding Protein
ALDH
Aldehyde Dehydrogenase
AR
Ankyrin Repeat
ARID1A
AT rich interactive domain 1A
bp
Base Pairs
BRCA1
Breast cancer 1, early onset
BRCT
BRCA1 C-Terminus
Brg-1
Brahama-related gene 1
BSA
Bovine Serum Albumin
C/EBPβ
CCAAT/Enhancer-Binding Protein β
COX
Cytochrome Oxidase
CREB
Cyclic AMP Response Element
CSC
Cancer Stem Cell
Cytc
Cytochrome c
DBD
DNA Binding Domain
DCIS
Ductal carcinoma in situ
DMEM
Dulbecco’s Modified Eagle Medium
DNA
Deoxyribonucleic Acid
Dox
Doxycycline
DTT
Dithiothreitol
EDTA
Ethylenediamine-Tetraacetic Acid
EGF
Epidermal Growth Factor
EMSA
Electrophoretic Mobility Shift Assay
EpiSC
Epiblast Stem Cell
x
ER
Estrogen Receptor
ERR
Estrogen Related Receptor
ESC
Embryonic Stem Cell
FBS
Fetal Bovine Serum
GABP
GA-Binding Protein
Gal
Galactose
Glu
Glucose
HAT
Histone Acetyl Transferase
HD
Heterodimerization Domain
HEPES
4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid
HIF-1α
Hypoxia Inducible Factor-1α
HRP
Horseradish Peroxidase
HSA
Helicase/SANT-associated
HSC
Hematopoietic Stem Cell
KCl
Potassium chloride
kDA
Kilodalton
LOH
Loss of Heterozygosity
LZ
Leucine Zipper
MaSC
Mammary Stem Cell
MgCl2
Magnesium Chloride
MOI
Multiplicity of Infection
mtDNA
mitochondrial DNA
Na2HPO4
Disodium Hydrogen Orthophosphate
Na3VO4
Sodium Orthovanadate
NaCl
Sodium Chloride
xi
NaF
Sodium Fluoride
NFAT
Nuclear Factor of Activated T-cells
NLS
Nuclear Localization Signal
NRF-1
Nuclear Respiratory Factor 1
PARP-1
Poly [ADP-ribose] polymerase 1
PBS
Phosphate Buffered Saline
PCR
Polymerase Chain Reaction
PGC-1
Peroxisome Proliferator-activated receptor Gamma Co-activator 1
PLB
Passive Lysis Buffer
PMSF
Phenyl Methyl Sulfonyl Fluoride
PPARγ
Peroxisome Proliferator-Activated Receptor-γ
PRC
PGC-1 Related Co-activator
Ptd
Pointed Domain
PVDF
Polyvinylidene difluoride
qRT-PCR
Quantitative real-time polymerase chain reaction
REDOX
Reduction-Oxidation
RNA
Ribonucleic Acid
ROS
Reactive Oxygen Species
RT
Room Temperature
SDS
Sodium Dodecyl Sulfate
SDS-PAGE
Sodium Dodecyl Sulfate-Polyacrylamine Gel Electrophoresis
shRNA
Short hairpin Ribonucleic Acid
SP1
Specificity Protein
SRC3
Steroid Receptor Co-activator
SWI/SNF
SWItch/Sucrose NonFermentable
xii
TAD
Transactivation Domain
TBE
Tris/Borate/EDTA
TBP
TATA-binding protein
TDLU
Terminal Ductal Lobular Unit
TEB
Terminal End Bud
TFAM
Mitochondrial Transcription Factor A
WCL
Whole Cell Lysates
WT
Wild-type
xiii
CHAPTER 1: Introduction and Literature Review
1.1 Breast Cancer
With approximately 22,000 new cases of breast cancer expected to arise in Canada in
2012, breast cancer represents the most frequently diagnosed cancer in women affecting nearly 1
in 10 during their lifetime, and is the second leading cause of cancer mortality [1]. Clinical
classification of breast cancer has typically involved analyses based on stage, histopathology, and
grade, but molecular classification using expression profiling has stratified breast cancer into six
biologically distinct subgroups, reflecting the heterogeneity of the disease. These subgroups differ
in their occurrence [2,3], prognosis [4] and clinical behavior [5], and consist of luminal A,
luminal B, basal-like/triple negative, claudin-low, HER2/ERBB2 over-expressing, and normallike breast cancers [4,6]. These molecular subgroups are also broadly conserved across ethnic
groups [7] and are apparent in ductal carcinoma in situ, the invasive carcinoma precursor [8],
supporting the view that unique molecular networks contribute to the initial development of these
intrinsic tumor subtypes. Additional breast cancer subtypes have been proposed by a recent study
identifying novel biological subgroups through the examination of somatic copy number
aberrations and gene expression profiling [9]. In this study ten novel subtypes were identified,
each characterized by a distinct molecular profile and prognosis. Further studies integrating
epigenetic and micro-RNA profiles with the currently characterized genomic architecture will
likely increase the number of breast cancer subgroups as additional molecular mechanisms are
potentially involved. Even with the current classification of tumor subtypes, there remains a
considerable underlying complexity, which is reflected in the heterogeneity of responses within a
tumor subtype to chemotherapy, hormonal, and targeted therapies [10]. In order to improve
current therapies, a better understanding of the etiology of these tumor subtypes is necessary. The
1
identified molecular tumor subgroups have been suggested to arise from the transformation of
distinct breast cell types that exist within a cellular hierarchy [10], with the molecular profile of
the subgroups proposed to reflect the cell of origin [11]. As the cellular organization apparent
during mammary carcinogenesis is derived from the normal mammary gland, an understanding of
the normal epithelial hierarchy could provide insight into the origin and development of breast
cancer subtypes.
1.2 The Human Mammary Gland
1.2.1 Structure and Development
The mammary gland is comprised of a branching network of ducts that end in clusters of
small ductules, known collectively as terminal ductal lobular units (TDLUs) [12]. Embedded
within the mammary fat pad, these TDLUs are the basic functional units of the breast and consist
of luminal epithelial and myoepithelial cells. The luminal cells can differentiate into both the
ductal cells that line the hollow breast ducts, and the milk-producing alveolar cells that arise
during pregnancy. In contrast, myoepithelial cells are specialized contractile units that form the
basal epithelial cell layer directly surrounding the luminal cells and direct cell polarity through
the secretion of the basement membrane (BM) which forms the exterior [6].
Development of the mammary gland occurs primarily postnatally and consists of two
functionally distinct growth phases: a ductal phase which generates the basic tissue architecture,
and a secretory phase responsible for milk production necessary during pregnancy [13,14]. The
ductal phase begins at birth, where a rudimentary ductal tree stems from the nipple and exists in a
quiescent state until puberty, after which increased hormonal levels initiate the formation of the
mature ductal network [13]. These ducts exist in a highly dynamic state, with the distal ends
undergoing branching morphogenesis and invading the surrounding tissue stroma [13]. The
2
mature glandular tissue is established when the ductal branching, which first develops centrally
during puberty, expands to the periphery of the mammary gland terminating in TDLUs. Although
extensive mammary development occurs from puberty to maturity, there are significant glandular
changes associated with pregnancy. In response to circulating pregnancy hormones there is
increased ductal branching terminating in new TDLUs, where the luminal epithelial cells
differentiate into the alveolar cells necessary for milk production and secretion [12]. During postpregnancy lactation, the secretory cells present within the mammary gland are maintained
through mechanical stimulation from suckling [15]. However following weaning, the mammary
gland undergoes postlactational involution that reverts the tissue to the mature tissue resting state.
This stage involves apoptosis and phagocytosis of the alveolar cells in addition to inactivation of
lactation-associated genes and activation of involution-associated genes, with nearly 80% of the
mammary epithelial cells undergoing programmed cell death [12,16].
1.2.2 Epithelial Hierarchy and Tumourigenesis
The existence of a cell hierarchy within the mammary gland is supported by early studies
characterizing bipotent mammary epithelial progenitor cells capable of luminal- and
myoepithelial-restricted differentiation [17,18]. Mammary stem cells (MaSC) exist at the apex of
this hierarchy, with the ability to both self-renew and generate epithelial progenitor cells that can
differentiate along the luminal or myoepithelial cell lineage. MaSC have been shown to reside in
the ductal branches of the human mammary gland where they are relatively quiescent, while the
bipotent progenitor cells that result from MaSC differentiation exist in the lobules and
demonstrate a higher proliferative activity [19]. It has been suggested that the entire lobule and
ductal structures of the human mammary epithelium can be derived from a single stem cell [20]
3
and it has been shown that a single mouse mammary stem cell is able to reconstitute an entire
functional mammary gland [21].
Transformed MaSC have been suggested as the initiating and driving force behind
tumourigenesis given the unique properties and regulation of stem cells. Whereas differentiated
somatic cells are highly proliferative and short lived with constant turn over, stem cells are
relatively quiescent and long lived which allows for the accumulation of oncogenic mutations
from exposures to common mutagens such as ionizing radiation, environmental chemicals,
viruses and inflammation [22-24]. The maintenance of stem cells in a quiescent cell cycle state
also permits accrual of DNA damage through reduced activation of checkpoint controls and DNA
damage response pathways, which prevents repair of damaged DNA or induction of apoptosis
[25]. Furthermore, the defects acquired by quiescent stem cells could be propagated through selfrenewing progeny and progenitors, resulting in a transformed cell pool. In addition to the inherent
properties of stem cells, the molecular pathways regulating stem cell renewal including the Wnt,
Notch and PTEN signally pathways are often deregulated in cancer [26].
1.2.3 The Cancer Stem Cell Theory
The cancer stem cell theory proposes that malignant tumors exist in a cellular hierarchy
similar to normal tissues and arise from a small number of transformed cells known as cancer
stem cells (CSC), with non-CSC being unable to initiate tumor formation [27]. CSCs were first
identified in the hematopoietic cell system, where hematopoietic stem and multi-/oligopotentprogenitors capable of indefinite proliferation are able to generate all types of mature blood cells,
but their transformation results in leukemogenesis [28]. Breast CSCs can be isolated from nontumorigenic cancer cells using a combination of cell surface markers (CD44/CD24) and aldehyde
dehydrogenase activity (ALDH), where the phenotype CD44+CD24-/low and ALDH1high is
4
characteristic of breast CSCs [29]. Identification of CSC surface markers and evidence for the
cancer stem cell hierarchy in the mammary gland comes from transplantation studies in mice. In a
study by Al-Hajj et al., cells were isolated from nine human breast carcinomas based on the
expression of CD44 and CD24 markers, which were heterogeneously expressed in tumor tissue
[30]. The tumorigenic potential was assessed following implantation of the isolated cells into the
cleared mammary fat pads of NOD/SCID mice, and notably only the CD44 + CD24- cell
population was able to initiate tumor formation. Furthermore, this tumor-initiating subpopulation
of breast cancer cells was able to regenerate the heterogeneity of the original tumor, producing
CD44+ CD24- and CD44+/- CD24+ cells, and undergo serial transplantation demonstrating a
capacity for self-renewal. Interestingly, basal-like breast tumors are enriched with cells displaying
the CD44+CD24- stem cell marker phenotype, which is associated with a less differentiated and
more invasive phenotype [31].
1.2.4 Luminal Progenitors as the Cell-of-Origin in Basal-Like Tumors
With basal- and luminal-like cancers representing the two most prominent subtypes of
breast cancer and are identified based on their similarities to the basal and luminal lineages in the
normal mammary gland [32], it was initially speculated that these breast tumor subtypes
originated from lineage-specific transformed cells [33,34]. However, there is increasing evidence
that luminal progenitor cells contribute to both basal- and luminal-like cancers, with gene
expression signatures of basal- and luminal-like cancers most similar to luminal progenitor cells
[35-37]. Although luminal cells have traditionally been described as less malignant in comparison
to basal-like cells [30], these luminal progenitors have recently been shown to be tumor initiating
and invasive, without de-differentiating or regressing to a basal cell state [38]. The tumor
suppressor BRCA1 (breast cancer 1, early onset) has been shown to be a regulator of luminal cell
5
differentiation in human breast cancer, where its absence results in expansion of the luminal
progenitor population and a characteristic basal-like tumor phenotype [36].
1.3 BRCA1 and Tumourigenesis
1.3.1 Familial and Sporadic Breast Cancers
Hereditary breast cancer is associated with a high risk of developing breast cancer with
the increased susceptibility resulting from an inherited germline mutation in one allele of a high
penetrance tumor suppressor gene including BRCA1, BRCA2, TP53, or PTEN [39]. Inactivation
of the second allele through loss of heterozygosity (LOH) is able to initiate the oncogenic
process, with this mechanism of tumourigenesis referred to as the two-hit hypothesis [40].
Additionally, BRCA1 haploinsufficiency results in a diminished DNA-repair capacity which has
been suggested to accelerate the development of hereditary breast carcinogenesis and facilitate
the accumulation of additional mutations [41]. Germline mutations in BRCA1 account for
approximately 30% of hereditary breast cancer cases, with hereditary cases representing only
about 10% of total breast cancer cases [42]. The remainder of breast cancer cases are sporadic in
nature, with only a limited number of cases attributed to mutations in BRCA1. However, reduced
BRCA1 mRNA levels are frequently observed in sporadic breast cancers, with decreased BRCA1
expression levels correlating with increased tumor aggression, enhanced cancer metastasis, and a
poorer clinical prognosis [43,44]. Therefore, an underlying defect in BRCA1 expression or in
related pathways will have the same functional consequence as mutational inactivation of BRCA1
itself. Methylation of the BRCA1 promoter has also been suggested as a contributing factor for
decreased BRCA1 expression in sporadic breast tumors [45]. However, decreased expression of
BRCA1 due to promoter methylation is observed in approximately 11% of sporadic breast cancer
tumors [45-47].
6
1.3.2 BRCA1 as a Tumor Suppressor
There is substantial experimental evidence indicating that BRCA1 functions as a tumor
suppressor. BRCA1 has been shown to be involved in many cell functions that serve to maintain
genomic integrity, and its disruption often results in defective G1/S, S, and G2/M cell checkpoints
[48-50]. In addition to interacting with proteins involved in cell cycle regulation, BRCA1
interacts with other tumor suppressors and oncogenes, proteins involved in DNA damage repair,
and transcriptional activators and repressors [51]. As inactivation of BRCA1 alone is not
sufficient for tumourigenesis, cell cycle checkpoint dysfunction coupled with impaired DNA
damage repair has been proposed to induce genetic instability in BRCA1 deficient cells,
increasing the risk for malignant transformation [52]. The progression from genetic instability to
tumor formation through oncogene activation is supported by the expression profile of BRCA1associated tumors, which are frequently associated with increased expression of oncogenes such
as cyclin D1, c-Myc and ErbB2 [53,54].
1.3.3 BRCA1 in Development and Differentiation
Studies of murine Brca1 expression during development have revealed that it is tightly
regulated at distinct stages of mammary gland development [55]. During early embryogenesis
(E9.5-10.5) Brca1 is diffusely expressed among the embryonic tissues, where it shifts in late
embryogenesis (E13-20) towards tissue specific expression in cells undergoing rapid growth and
differentiation [56]. Postnatal expression of Brca1 in the breast remains at low levels until
puberty, where a sudden increase in expression accompanies ductal morphogenesis in the
mammary gland with the formation of proliferative terminal end buds (TEB) [55]. In the adult
tissue Brca1 expression is localized primarily in proliferating tissues undergoing differentiation,
with expression in the mammary gland remaining low in the virgin mouse until pregnancy where
7
a dramatic increase in expression accompanies the formation of milk-secreting alveolar units in
the TEB [56]. With Brca1 expression remaining elevated during pregnancy and lactation, postlactational regression stimulated by weaning results in decreased Brca1 expression. A final
increase in Brca1 expression accompanies involution of the mammary gland, which is triggered
by a massive apoptotic event, and returns the mammary gland to a quiescent state [55]. Although
this quiescent state is associated with low Brca1 expression, this baseline expression is higher
than in age-matched virgin mice [56]. These increased levels of Brca1 support human
epidemiological studies describing the protective effect of pregnancy in breast cancer
development for parous women [57].
Several studies have examined the crucial role of Brca1 in development using targeted
gene knockout. Mouse models with global homozygous deletions in Brca1 revealed an early
embryonic lethality (E5.5-13.5), supporting a role for Brca1 in embryonic development. These
embryos exhibited various phenotypic deficiencies including developmental delays and structural
chromosomal aberrations which were likely due to accumulated DNA damage [58], consistent
with a role for BRCA1 in the maintenance of genome integrity. Similar to BRCA1 knockouts,
truncations of BRCA1 eliminating the BRCT domain, which has recently been shown to be
required for its tumor suppressor function [59], results in a delayed early embryonic lethality
(E9.5-10.5). Embryogenesis was associated with a delay in development and growth arrest but
continual proliferation and differentiation [60]. Therefore, truncations appear to have a different
effect on embryonic development than complete protein absence. To overcome the embryonic
lethality phenotype associated with BRCA1 knockout, Xu et al. utilized conditional knockout of
BRCA1 in the mammary gland using a Cre-loxP system [61]. Abnormal ductal morphogenesis
and increased apoptosis was observed following conditional knockout, with 80% of mice
8
remaining tumor free until 10 months of age. Notably, the low frequency and delay in tumor
formation following conditional Brca1 knockout in these mice suggests that additional
transformations combined with loss of Brca1 are required for tumourigenesis.
The requirement for BRCA1 in embryogenesis and mammary gland development
suggests it plays a regulatory role during cell proliferation and differentiation. Using murine
mammary epithelial cells that differentiate in vitro in response to lactogenic hormones, BRCA1
has been shown to promote differentiation while down-regulating proliferation assessed by the
expression of the milk protein β-casein and formation of domes [62]. Additionally, its depletion
was shown to result in attenuation of differentiation. A similar study depleted BRCA1 using RNA
interference in the human mammary epithelial cell line MCF-10A, which resulted in the
formation of abnormal cell aggregates having increased cell proliferation during differentiation
[63]. BRCA1-depleted cells were also associated with down-regulation of genes involved in
differentiation and up-regulation of genes involved in proliferation. As mammary gland
development is based on the differentiation of stem/progenitor cells along the luminal or
myoepithelial lineage, the influence of BRCA1 expression on the epithelial hierarchy has been
previously investigated [64]. Lim et al. examined the breast tissue of BRCA1 mutation carriers
(BRCA1+/-) and observed a diminished stem and enriched luminal progenitor expression signature
[36]. In contrast, Liu et al. demonstrated that while BRCA1 expression increases in vitro during
mammary cell differentiation, knockdown of BRCA1 (~20% expression) results in expansion of
the undifferentiated stem/basal cell population with a corresponding decrease in luminal cells
[64]. These results suggest that the effects of BRCA1 on mammary stem/progenitor
differentiation may be dose dependent. BRCA1 heterozygosity results in expansion of the luminal
cell population while homozygous BRCA1 inactivation shifts the cell population towards a
9
stem/basal-like state. Interestingly, this model is consistent with BRCA1 expression in normal
cells where it remains low in the stem/basal population, while being nearly 3-fold higher in
luminal cells [64]. Collectively, these findings establish BRCA1 as an important regulator of cell
growth and differentiation.
1.4 Regulation of the BRCA1 Gene
Characterization of the transcription factors that regulate the BRCA1 proximal promoter
has previously been performed with several protein complexes identified [55]. The GA-binding
protein (GABP) is a critical regulator of promoter activity and has been shown to bind the RIBS
site [45]. Recently, the low BRCA1 expression present in the human mammary carcinoma, ErbB2
amplified, SK-BR-3 cell line was attributed to decreased expression of the GABP subunit,
GABPβ, that resulted from aberrant regulation of a site within the GABPβ promoter by a multiprotein complex containing Nuclear Respiratory Factor-1 (NRF-1) [65]. As aberrant activity of
NRF-1 or its regulatory complex members are able to modulate BRCA1 expression through their
regulation of GABPβ, further studies are required to identify these complex co-activators. GABP
also mediates the interaction of the unliganded glucocorticoid receptor with the BRCA1 promoter
at the RIBS site in the absence of the stress hormone hydrocortisone [66], supporting a
mechanism in which stress regulates BRCA1 expression (reviewed in [67]). A CREB (Cyclic
AMP Response Element Binding protein) site exists downstream of the RIBS element and acts as
a constitutive transcriptional activation element bound by CREB [68]. It was subsequently
demonstrated that c-Jun/Fra2 also bind this site and positively regulate BRCA1 expression [69].
Positive regulation has also been demonstrated by a complex containing 53BP1 (p53 binding
protein) at the UP site [70]. The oncogene Steroid Receptor Co-activator 3(SRC3) has been
demonstrated to complex with 53BP1 to regulate the BRCA1 promoter, possibly at the UP site
10
[71]. Although the mechanism by which p53 transcriptionally represses BRCA1 expression is not
known [72], it is possible that its interaction with bound 53BP1 suppresses its activity and
reduces BRCA1 expression. Regulatory control of the promoter also appears to be mediated by
several E2Fs, where binding of either E2F6 or E2F1 has been shown to repress and activate the
promoter, respectively [73]. Additionally, bound E2F1 and E2F4 at adjacent E2F binding sites
results in negative regulation of the promoter [74]. Similarly, repression of the BRCA1 gene is
mediated upon binding of Rb to E2F1 at the E2F binding site [75]. BRCA1 expression is also
shown to increase in response to treatment with estrogen [76]. Its interaction with estrogen is
complex as BRCA1 is able to both repress ERα signalling and stimulate its expression (reviewed
in [77]). The involvement of BRCA1 in hormone signalling may contribute to the associated
breast- and ovarian-specific tumor development associated with BRCA1 mutation in hereditary
[78] and BRCA1 inactivation in sporadic cancers [47,79,80]. In addition to regulation of the
proximal promoter, a recent study has demonstrated c-Myc activation of the promoter through
distal regulatory elements [81].
Regulation of BRCA1 transcripts has recently been shown to involve micro RNAs
(miRNA). The decreased expression of BRCA1 associated with sporadic breast cancers has been
suggested to be mediated by miR-182 in the 3’UTR of BRCA1 [82], resulting in diminished
DNA damage response capacity. Additionally, miR-335 has been shown to modulate the BRCA1
regulatory cascade in vitro through regulation of the BRCA1 activators ERα, IGF1R, SP1, and
the repressor ID4 with miR-335 expression itself being regulated by estrogen [83]. Upstream
deregulation of miRNAs impacting BRCA1 can result in an amplified downstream effect given
recent findings that BRCA1 epigenetically controls the oncogenic miR-155 [84] with targets
involved in apoptosis and proliferation, both of which contribute to tumor development.
11
1.5 GAPB Regulation and Cellular Functions
GABP is a critical regulator of the BRCA1 promoter through the RIBS element [45].
GABP is a unique ets transcription factor that exists as an obligate multi-subunit protein complex
[85]. It consists of a GABPα and GABPβ heterodimer, which can in turn form a transcriptionally
active tetrameric α2β2 complex [85]. GABPα contains the DNA-binding domain (DBD) which is
characterized by the winged-helix-turn-helix motif common to ets transcription factors, and is the
only ets factor that can recruit GABPβ to DNA [86]. Its central region also contains an ets
pointed domain (Ptd), which mediates additional protein-protein interactions [87]. GABPβ
contains ankyrin repeats (AR) that form the interface with GABPα, a transactivation domain
(TAD) required for transcriptional activation and a leucine zipper (LZ) motif that mediates
tetramerization [85]. In humans there are four different GABPβ splice variants that differ in their
C-termini [88], and are named according to their molecular weight [85]. These GABPβ isoforms
all contain an N-terminal AR domain and a TAD, but differ in their inclusion of a central 12amino acid region (In) having no known function and a C-terminal LZ domain [85]. The structure
of human GABPα and the GABPβ isoforms are summarized in Figure 1.1. There is currently no
consensus on the precise location of the TAD and transcriptional activity of these isoforms.
Gugneja et al. reported that all four GABPβ isoforms exhibit indistinguishable transcriptional
properties with the TAD located directly upstream from the LZ domain [88]. In contrast, Sawa et
al. localized the TAD to the extreme C-terminal region, with only GABPβ1-42 and and
GABPβ1-41exhibiting transcriptional activation properties [89]. Furthermore, Rosmarin et al.
utilized in vitro luciferase assays which revealed that although all isoforms were capable of
promoter activation, the GABPβ1-42 and GABPβ1-41 isoforms show greater transcriptional
activation [85]. The
GABP
complex
is
12
involved
in
a
variety
of
cellular
A.
Ptd
DBD
HD
GABPα
1
154
318
203
454
399
B.
AR
In
TAD
LZ
GABPβ1-42
1
130
1
130
1
130
1
130
195 206
270
346
395
GABPβ1-41
258
334
383
GABPβ1-38
195 206
270
346 359
GABPβ1-37
C.
HD
258
NLS
334 347
DBD
TAD
TAD
NRF-1
1
78
88
116
301 304
380
449
476 503
Figure 1.1. Domain structures of GABPα, GABPβ1 isoforms and NRF-1. (A) The functional
domains of the GABPα subunit include an ets pointed domain (Ptd, dark grey), a helix-loop-helix
DNA-binding domain (DBD, green) and a heterodimerization domain (HD, orange). (B) Four
GABPβ1 splice variants have been identified, with their names derived from their molecular
weight. The functional domains of the GABPβ1 subunits include an ankyrin repeat
heterodimerization motif (AR, dark grey) and a transcriptional activation domain which contains
the nuclear localization signal (TAD, green). Isomer β1-42 and β1-41 also contain a C-terminal
leucine zipper domain (LZ, yellow), while β1-42 and β1-38 additionally possess a 12 amino acid
insert of unknown function (In, blue). (C) The functional domains of NRF-1 include a
homodimerization domain (HD, dark grey), a nuclear localization signal (NLS, blue), a large
DNA-binding domain (DBD, green), and a bipartite transcriptional activation domain (TAD,
orange).
13
functions including regulation of mitochondrial cellular respiration, cell-cycle progression and
differentiation [90,91]. In addition to mitochondrial metabolism which will be discussed below,
GABPα has recently been demonstrated to be crucial for the maintenance and differentiation of
hematopoietic stem/progenitor cells (HSC) [92]. Transcriptome analysis revealed that direct
GABP target genes included the transcription factors Zfx and Etv6, and prosurvival Bcl-2 family
members, all of which are important for HSC survival [92]. Maintenance of HSC quiescence was
also linked to GABP through its direct regulation of Foxo3 and Pten [92]. Additionally, GABP is
involved in the regulation of HSC proliferation and differentiation through transcriptional
activation of DNA methyltransferases and histone acetylases [92]. Furthermore, GABPα has also
been implicated in the regulation of Oct-3/4 expression, a factor necessary for the self-renewal of
embryonic stem cells [93]. Therefore, consistent with its regulation of BRCA1, it is likely that
GABP is involved in the regulation of breast differentiation.
A crucial regulatory role for GABP during development has been demonstrated in vivo
with mutational inactivation of GABPα [94]. Homozygous GABPα-/- mice exhibit a lethal
phenotype prior to implantation in the uterine endometrium. While heterozygous mice are not
phenotypically different from wild-type mice, they interestingly do not show a reduction in
GABPα protein levels suggesting tight regulation of its expression. The early embryonic lethality
is suggested to result from mitochondrial and nuclear target gene deficiencies. Notably, an
embryonic lethality phenotype is also observed with several GABP target genes including
mitochondrial transcription factor A (Tfam) [95] and BRCA1 [58].
1.6 Nuclear Regulators of Mitochondrial Metabolism
The mitochondrial genome has a limited coding capacity, with the 16.5 kb of closed
circular DNA encoding only 13 polypeptides that function as respiratory subunits of complexes I,
14
III, IV, and V, 22 mitochondrial transfer RNAs (tRNA) and 2 mitochondrial ribosomal RNAs
(rRNA) [96]. As approximately 100 different components are required for the mitochondrial
oxidative phosphorylation system alone, the nuclear genome is required to encode the remaining
components involved in oxidative phosphorylation, mitochondrial DNA transcription, translation
and replication, HEME biosynthesis and protein import and assembly [96,97]. These nuclearmitochondrial interactions are regulated by the family of Peroxisome Proliferator-activated
receptor Gamma Co-activators (PGC)-1 which includes PGC-1α, PGC-1β, and PGC-1 Related
Co-activator (PRC) [98]. Transcriptional regulation of nuclear encoded mitochondrial proteins by
this family of nuclear co-activators requires their recruitment to gene promoters through
interactions with DNA-binding transcription factors including NRF-1, GABP, Estrogen-Related
Receptors (ERR), and CREB, among others [99]. Finally, additional recruitment of histone acetyl
transferase (HAT) containing co-activator proteins and SWI/SNF chromatin remodeling
complexes is required to potentiate the transcriptional activity [99]. The regulatory involvement
of GABP and NRF-1 in mitochondrial biogenesis is summarized in Figure 1.2.
1.6.1 GABP and NRF-1 as Regulators of Mitochondrial Biogenesis
A number of GABP binding sites are located within the promoters of genes involved in
oxidative phosphorylation, mitochondrial import, and mitochondrial transcription factors
involved in mitochondrial genome replication [97]. Consistent with its role in oxidative
phosphorylation, knockdown of GABPα expression is sufficient to reduce expression of all 10
nuclear-encoded cytochrome c oxidase (COX) subunit genes in addition to genes required for
mitochondrial genome replication [100]. The ability of GABP to regulate gene expression is
controlled by the reduction-oxidation (redox) status of the mitochondria, with the DNA-binding
15
Figure 1.2. NRF-1 and GABP coordinately regulate the expression of nuclear-encoded
mitochondrial proteins (Adapted from Scarpulla, Physiol. Rev 88: 611-638, 2008). Nuclear
co-regulators (PGC-1α, PGC-1β and PRC) interact with DNA-binding transcription factors
including NRF-1 and GABP to regulate the expression of nuclear genes required for
mitochondrial maintenance and function. Their target genes have distinct functions in protein
import and assembly, heme biosynthesis, oxidative phosphorylation through their expression of
the majority of respiratory subunits, mitochondrial translation, and mitochondrial DNA (mtDNA)
transcription and replication. Abbreviations used: NRF-1, Nuclear Respiratory Factor-1; GABP,
GA-Binding Protein; PGC-1α, PPARγ co-activator-1 alpha, PGC-1β, PPARγ co-activator-1 beta;
and PRC, PGC-1 related co-activator.
16
ability of the GABPα subunit shown to be severely reduced in conditions of oxidative stress
[101].
NRF-1 was first identified as a regulator of the rat cytochrome c gene, but binding sites
were subsequently identified in several respiratory protein promoters (reviewed in [90]). NRF-1
binds to its recognition sequences as a homodimer, where its carboxy-terminal domain is utilized
for transcriptional activation and phosphorylation of its amino-terminal domain enhances its
DNA binding ability [102] (Figure 1.1c). A chicken homologue of NRF-1 functions as a
transcriptional repressor following glycosylation [103], suggesting that post-translational
modifications might serve to regulate NRF-1 function. A fundamental role for NRF-1 in
mitochondrial biogenesis is demonstrated by in vitro studies under conditions of NRF-1 depletion
[104]. Mitochondrial biogenesis in mouse myoblasts can be stimulated by PGC-1α, however a
dominant-negative NRF-1 is able to completely abrogate the effect of PGC-1α. Furthermore,
transactivation of the mitochondrial transcription factor A (Tfam) promoter by PGC-1α requires
interactions with NRF-1 and GABP [104]. Mutation of the NRF-1 binding site completely
abolishes activation of the promoter by PGC-1α, while the normal activity is reduced by a much
smaller extent in the presence of a mutated GABP binding site. These results collectively suggest
that NRF-1 is a primary target and an essential component of the PGC-1α regulatory network.
1.6.2 NRF-1 during Development
Similar to GABP and BRCA1, the requirement for NRF-1 expression during early
embryogenesis has been demonstrated with in vivo targeted disruptions in mice. A periimplantation lethal phenotype results from homozygous NRF-1 knockdown between embryonic
days 3.5-6.5, and is associated with reduced mitochondrial DNA content [105]. Although
heterozygous NRF-1+/- knockout mice are viable, they have a reduced birth weight and a higher
17
mortality with diminished mitochondrial activity. However, the embryonic lethality phenotype
observed with NRF-1 null mice is not attributed solely to reduced mitochondrial DNA content as
Tfam knockout mice, although embryonic lethal, survive until embryonic days 8.5-10.5 [95]. The
difference in embryonic viability for NRF-1 was attributed to its non-mitochondrial regulatory
functions and potentially disruptions of genes involved in cell growth [90].
1.7 Mitochondrial Metabolism in Breast Cancer
A unique feature of tumor cells is their preference to metabolize glucose by aerobic
glycolysis, which is in contrast to the mitochondrial oxidative phosphorylation performed by
normal cells [106]. Although both tumor and rapidly proliferating normal cells are characterized
by an increase in glucose consumption and glycolytic activity, tumor cells produce lactic acid
rather than metabolizing glucose in the citric acid cycle, a phenomenon first identified by Otto
Warburg more than 75 years ago [106-108]. Although the generation of ATP through aerobic
glycolysis is inefficient, it facilitates the uptake and incorporation of nutrients into the biomass
needed for rapid cell proliferation [109]. Although the mechanism for the metabolic shift is
unknown, there are two predominant theories. Warburg initially proposed that the shift towards
aerobic glycolysis was the result of impaired mitochondrial function [107]. Although mutations in
mtDNA are frequently observed in human cancers (reviewed in [110]) and in vitro defects in
oxidative phosphorylation have been shown to activate the oncogene Akt [111], the prevalence of
impaired oxidative phosphorylation in breast tumors is varied. Therefore, further studies are
required to assess the contribution of impaired mitochondrial function to tumourigenesis. In
contrast, activation of oncogenes including MYC and HIF has been suggested to contribute to the
Warburg phenotype through activation of glycolytic genes, which are suggested to shift
metabolism towards glycolysis (summarized in [106]). As tumor cell metabolism is complex,
18
further studies are required to investigate the importance of mitochondrial function in
tumourigenesis.
1.8 Rationale and Hypothesis
An understanding of the pathophysiological mechanisms involved in the initiation and
progression of breast cancer is necessary for the identification of therapeutic targets to be used in
its treatment. As the NRF-1 containing complex that regulates the GABPβ promoter is a positive
regulator of BRCA1 expression, the downregulation of BRCA1 frequently observed in sporadic
breast cancers could be mediated by decreased GABPβ, NRF-1 or other unidentified complex coactivators. Consistent with the fundamental role for BRCA1 in mammary epithelial
differentiation and recent evidence for the involvement of GABP in the differentiation of
stem/progenitor cells, NRF-1 might also play a role in cellular differentiation. Given the
important regulatory functions of NRF-1 and GABP in maintaining mitochondrial metabolism,
disruption of their expression could also result in the altered tumor metabolism frequently
observed during breast carcinogenesis and provide a link between the impaired differentiation and
altered mitochondrial metabolism associated with breast cancer. Therefore, the aim of this study
was to identify the components of the NRF-1 containing complex that regulates the GABPβ
promoter and to investigate the role of NRF-1 during BRCA1-mediated differentiation of
pluripotent breast cell lines.
19
CHAPTER 2: Isolation of a complex containing NRF-1 that regulates the GABPβ promoter
2.1 Abstract
Background: Sporadic cases of breast cancer are associated with decreased BRCA1 expression,
which is proposed to contribute to tumor progression. A critical regulator of the BRCA1 promoter
is GABP, a heterotetramer of GABPα and GABPβ. A regulatory complex containing NRF-1 has
been previously shown to activate the GABPβ promoter, where its decreased activity results in
low BRCA1 expression through downregulation of GABPβ promoter activity. As the
transcriptional regulation of the GABPβ promoter by the complex containing NRF-1 has not been
fully elucidated, we attempted to identify the additional co-activators of the complex.
Results: In the present study, we utilized an immobilized-template assay with a multimerized
GABPβ promoter binding site for the regulatory complex containing NRF-1, to isolate and
identify the complex co-activators. Competition assays revealed the presence of non-specific
DNA-binding proteins in the assay eluate. While optimization of binding and elution conditions
through variations of Mg2+ content and pH minimized binding of non-specific proteins, they were
still observed to elute from the immobilized-template with NRF-1 and presumably the coactivators. Furthermore, complex formation at the GABPβ promoter with the isolated fractions in
an EMSA was not consistent with the elution profile of NRF-1.
Conclusions: The presence of non-specific DNA-binding proteins in the immobilized-template
eluate prevented the identification of co-activators in the NRF-1 containing complex that
regulates GABPβ activity. Modifications to the assay would be required to conclusively identify
members of the regulatory complex.
20
2.2 Background
Women with germline mutations in the tumor-suppressor gene BRCA1 (breast cancer 1,
early onset) are predisposed to high lifetime risks of developing breast and ovarian cancers, with
penetrance values of approximately 65% for breast and 39% for ovarian cancers by age 70
[112,113]. Hereditary breast cancers account for 5-10% of all breast cancer cases, with only 30%
of these cases being attributed to germline mutations in BRCA1 [42,114,115]. The non-inherited
sporadic forms of breast cancer make up the remainder of breast cancer cases, with few being
associated with mutations in BRCA1 [55,116]. However, reduced BRCA1 mRNA levels are often
observed in sporadic breast cancers, with decreased BRCA1 expression levels correlating with
increased tumor grade, cancer metastasis, and a poorer clinical prognosis [43,44,117]. As only a
small proportion of breast cancer tumors present with BRCA1 promoter hypermethylation,
decreased BRCA1 expression through altered transcriptional regulation has been suggested as
contributing to sporadic breast cancer progression [45,46,118]. Therefore, a focus should be
placed on evaluating the BRCA1 regulatory network to determine the basis for the decreased
BRCA1 expression observed in sporadic breast tumors.
BRCA1 transcriptional regulation is complex and under the control of a highly active
bidirectional promoter regulated by DNA-binding proteins, transcriptional co-activators and corepressors [55,119]. One critical activator of BRCA1 expression is the GABP DNA-binding
complex, which binds the RIBS element of the BRCA1 promoter [45]. The ets transcription factor
GA-binding protein (GABP) is transcriptionally active following heterotetramer formation,
consisting of two GABPα and GABPβ subunits, which contain the DNA-binding domain and
transcriptional activation domain, respectively [85,86,120]. Recently, it has been demonstrated
that the low BRCA1 expression characteristic of the Erb-B2-amplified SK-BR-3 cell line is the
21
result of reduced GABPβ promoter activity and involves a multi-protein complex containing
Nuclear Respiratory Factor 1 (NRF-1), that binds and activates the promoter approximately 270
bp upstream from the transcription start site [65]. The decreased GABPβ expression was not
attributed to altered NRF-1 expression, but instead to aberrant binding of a complex co-activator.
To elucidate the mechanism behind the altered GABPβ promoter activation and subsequent
down-regulation of BRCA1, identification of the complex co-activators is required to understand
the basis for their aberrant promoter binding and activation.
In this study, we set out to identify the unknown co-activators that comprise the GABPβ
regulatory complex containing NRF-1. To this end, we developed an immobilized-template
affinity assay similar to that which has proven useful in the identification of brahma-related gene
1 (Brg-1) and Specificity Protein 1 (Sp1) in the regulation of the SPARC promoter [121]. This
assay allowed for the capture and purification of proteins from nuclear extracts, which were
demonstrated to bind the GABPβ promoter at the site activated by the complex containing NRF1. Assay conditions were optimized to minimize the purification of non-specific DNA binding
proteins, which is a frequent occurrence when proteins are isolated by DNA affinity. Here, we
described the isolation of a complex containing NRF-1 that binds the GABPβ promoter. We also
reported that in addition to the complex isolation, there were several non-specific binding proteins
present, which made identification of the specific co-activators in the complex containing NRF-1
through mass spectrometry unattainable. A discussion on purification strategies to circumvent the
presence of non-specific DNA-binding proteins is also presented.
22
2.3 Methods
2.3.1 Cells and cell culture
The human mammary epithelial adenocarcinoma cell line, MCF-7, was obtained from the ATCC
(Manassas, VA, USA). MCF-7 cells were maintained in RPMI 1640 medium supplemented with
10% fetal bovine serum, 100 units/ml penicillin, and 100 g/ml streptomycin, humidified at 37oC
with 5% CO2.
2.3.2 Design of dual biotin-labelled probes
The GABPβ promoter sequence spanning nucleotides -268/-251 bp upstream from the
transcription start site was repeated in triplicate following an AGCT overhang to which a dual
biotin label was bound at the 5’ end for Gb-270multiBT. The same region of the promoter was
repeated in triplicate following a dual biotin-labeled AGCT with conserved mutations of residues
-252/-251 for Gb-270(m10)multiBT. The mutated residues were predicted by TRANSFAC
Alibaba 2.1 to abolish Sp1 binding to the probe. Full double stranded sequences of the probes are
given in Figure 2-1.
2.3.3 Immobilized-template assays
Immobilized-template assays were performed as described previously [121]. Two hundred
micrograms of Dynabeads M280 streptavidin (Sigma Aldrich) was concentrated using a magnet
and resuspended in 100 μl of buffer T (10 mM Tris [pH 7.5], 1 mM EDTA, 1 M NaCl). The
beads were again concentrated and resuspended in buffer T for a total of three washes. The beads
were then concentrated and resuspended in 16 μl of buffer T with 10 pmol of Gb-270multiBT or
probe Gb-270(m10)multiBT. The mixture was agitated by flicking for 1 hr at room temperature
(RT) and the beads were washed 3 times in 100 μl of buffer T to remove unbound probes. Beadcoupled probes were equilibrated in buffer R (10 mM Tris pH 7.5, 1 mM MgCl2, 0.1% IgePal, 1
23
mM EDTA, 10 mM DTT, 5% glycerol, 60 mM KCl, 12 mM HEPES pH 7.9, 0.03% BSA), buffer
Z (25 mM HEPES pH 7.6, 60 mM KCl, 12.5 mM MgCl2, 1 mM DTT, 5% glycerol, 0.1% IgePal,
0.03% BSA), buffer Ze (25 mM HEPES pH7.6, 60 mM KCl, 1 mM DTT, 5% glycerol, 0.1%
IgePal, 0.03% BSA), or buffer Ze (pH 7.0, 7.5, 8.0) for 30 min RT. Beads were centrifuged,
isolated and resuspended in buffer R/Z/Ze, to which 200 μg MCF-7 nuclear extracts (prepared as
described previously [65]) and 40 ng/μl poly (dG-dC) were added (120 μl final volume) and
mixed for 30 min by flicking at 4oC. Beads were collected and unbound proteins were removed
and saved. The beads were washed three times with buffer R/Z/Ze containing 10 ng/μl poly (dGdC). Bound proteins were either immediately eluted by boiling in 1X SDS loading buffer, or
through gradient salt elution using buffer R/Z/Ze containing 100-1000 mM KCl with two washes
performed and pooled for each salt concentration. The presence of ARID1A, Brg-1, Sp1, TBP,
and NRF-1 were detected by Western blot analysis. Large scale immobilized-template assays
were scaled up from the normal protocol by a factor of 5, utilizing 1 mg MCF-7 nuclear extracts
with 600 μl binding and elution volume.
2.3.4 Immobilized-template competition assay
MCF-7 nuclear extracts (100 μg) were pre-incubated with 50 pmol free NFY, Gb-270, Gb270multiBT, or m10 competitors in a final volume of 50 μl adjusted with buffer R, and gently
agitated on ice for 30 min. Nuclear extracts with and without competitors were added to
immobilized Gb-270multiBT or Gb-270(m10)multi BT (outlined above) which had previously
been equilibrated with buffer R for 30 min RT. Beads with nuclear extracts and competitors were
agitated for 30 min RT, after which beads were collected and unbound proteins isolated. Beads
were washed three times with buffer R containing 10 ng/μl poly (dG-dC), and bound proteins
24
were eluted by boiling in 1X SDS loading buffer (2.5% sodium dodecyl sulfate, 25mM Tris-HCl
pH 6.8, 100mM DTT, 10% glycerol).
2.3.5 Western blot analysis
Isolated fractions were resolved on a SDS-polyacrylamide gel, transferred to PVDF membrane,
and probed with the appropriate antibody. Primary antibodies included: anti-ARID1A (2035C5a,
1:500, Abcam), anti-Brg-1(H-88, 1:1000, Abcam), anti-Sp1 (H-225, Santa Cruz Biotechnology),
anti-TBP (1TBP18, 1:2000, Abcam) and anti-NRF-1 (M01, 1:250, Abnova). Secondary antibody
detection was performed using SuperSignal West Pico Chemiluminescence Substrate (Thermo
Scientific/Fisher, Nepean, Canada).
2.3.6 Electrophoretic mobility shift assay (EMSA)
Nuclear extracts (5 μg) or isolated fractions (2 μl of 120 μl volume) were combined with
32
P-
labelled oligonucleotides (1 ng) in binding buffer (25 mM HEPES pH 7.6, 5 mM MgCl 2, 34 mM
KCl, 50 μg/ml poly (dI-dC) (Sigma), 0.5 mg/ml bovine serum albumin (BSA)) and incubated on
ice for 15 min. Samples were resolved on a 6% acrylamide 0.25 x TBE non-denaturing gel, after
which the gel was fixed (20% ethanol, 10% acetic acid) and dried onto filter paper. The Gb-270,
Gb-270m6 and RC4 oligonucleotide sequences have been previously reported in [65].
2.4 Results
2.4.1 Isolation of the NRF-1 containing protein complex that binds the GABPβ promoter using an
immobilized-template assay
A multi-protein complex containing NRF-1 has previously been shown to regulate the
GABPβ proximal promoter, -268/-251 bp upstream of the transcription start site [65]. To identify
additional proteins in this complex that bind -268/-251, an immobilized template affinity assay
was employed. The -268/-251 sequence of the promoter was multimerized to increase capture of
25
the protein complex, and dual biotin-tagged to allow for immobilization on streptavidin-coated
magnetic beads (Figure 2-1a). TRANSFAC AliBaba 2.1 transcription factor prediction software
was utilized to identify potential artificial binding sites created by multimerizing -268/-251, and
identified a restored Sp1 binding site from two native binding sites present at either end of -268/251 (Figure 2-1a). Several nucleotides not required for binding by the GABPβ regulatory
complex, but key for the artificial Sp1 binding site in the multimer, were mutated to abolish Sp1
binding which was confirmed with prediction software. This mutated -268/-251 multimer was
dual biotin-tagged for an additional immobilized-template predicted to have reduced non-specific
protein binding (Figure 2-1b). Previous unpublished work using a maltose binding protein-fused
NRF-1 affinity column and EMSA identified the SWI/SNF proteins Brg-1 and ARID1A (AT rich
interactive domain 1A) as potential members of the -268/-251 regulatory complex. Binding of
Brg-1, ARID1A, NRF-1 and Sp1 to the multimerized -268/-251 was substantiated with Western
blot analysis of the immobilized-template assay column eluate in the presence and absence of
excess free -268/-251 monomer and multimer competitors to assess binding specificity (Figure
2.2). Although binding to the immobilized -268/-251 multimer by Brg-1 and ARID1A and NRF-1
could be demonstrated, only NRF-1 could be confirmed as binding specifically. While the
majority of NRF-1 was retained by the multimer, there was substantial unbound Brg-1 and
ARID1A. Also, free monomer and multimer was able to compete for NRF-1 binding resulting in
reduced recovery from the immobilized template, with no substantial effect on Brg-1 and
ARID1A. Therefore, the observed association of Brg-1 and ARID1A with the -268/-251 multimer
is likely the result of non-specific binding. Additionally, Sp1 was observed to bind the -268/-251
26
A.
NRF-1
Sp1
B.
NRF-1
Figure 2-1: Dual biotin-tagged probe for DNA affinity assay. Schematic representation of the 66bp immobilized template containing the
multimerized GABPβ promoter sequence, with a dual biotin (BT) tag at the 5’ end. The nucleotide sequence of the GABPβ promoter located -268
to -251bp (Gb-270) upstream of the transcription start site is enclosed by solid boxes and repeated in triplicate for Gb-270multiBT (A) with
predicted NRF-1 and Sp1 binding site indicated with dashed and solid lines, respectively. Gb-270(m10)multiBT (B) is derived from (A) with
dashed boxes indicating mutated residues in Gb-270, predicted to abolish non-specific Sp1 binding to the probe. Binding site predictions were
obtained from sequence analysis using TRANSFAC AliBaba 2.1.
27
Figure 2.2. Proteins bind non-specifically to the GABPβ promoter in an immobilizedtemplate assay. Nuclear extracts were prepared from MCF-7 cells. Proteins which bind to Gb270multiBT (Elution) were isolated from the nuclear extracts (Unbound) using the immobilizedtemplate assay and subjected to Western blot analysis with antibodies against ARID1A, Brg-1,
SP-1 and NRF-1, with non-specific binding to beads assessed (Beads Only). Specificity of
binding was determined by pre-incubating proteins with free competitors in separate assays prior
to addition to the immobilized-template. Competitors included NFY probe (NFY) lacking the
binding site, Gb-270 (Gb-270), or Gb-270(m10) (m10).
28
multimer in the presence and absence of competitors suggesting its association is also nonspecific, supporting the need for a mutated -268/-251 multimer to reduce non-specific binding.
2.4.2 Non-specific binding proteins elute with NRF-1 from the immobilized-template
In an attempt to isolate NRF-1 and its associated regulatory complex from proteins
binding the -268/-251 multimer non-specifically, a gradient salt elution of bound proteins was
performed to preferentially disrupt binding of weaker binding non-specific proteins using low
salt, while maintaining complex binding which should dissociate from the template at higher salt
concentrations. The gradient salt elution was performed with the -268/-251 multimer and mutated
multimer (Figure 2.3). Although NRF-1 was observed to elute between 300-400 mM KCl with
both immobilized templates, there was considerable overlap with the non-specific binding Brg-1,
ARID1A and Sp1. There was a general reduction in protein binding with the mutant -268/-251
multimer, but this included both non-specific proteins and NRF-1. The binding of Sp1 to both the
-268/-251 multimer and mutated multimer suggests that either the mutated residues of the
multimer were not sufficient to disrupt DNA binding at the artificial binding site, or that Sp1 is
binding non-specifically with Brg-1 and ARID1A to the DNA or bound proteins potentially
mediated by the binding and elution conditions.
2.4.3 Optimization of the immobilized-template binding and elution conditions
The effects of divalent metal ion cofactors such as magnesium on protein-DNA
interactions are varied, with its presence observed to enhance sequence specific binding [122],
weaken protein-DNA interactions, or leave them unaffected [123]. Immobilized template assays
were therefore conducted with the -268/-251 multimer in the presence or absence of magnesium
in the binding and elution buffers, or a combination of binding in the presence of magnesium and
elution without it. Following gradient salt elution of bound proteins, there was generally an
29
A.
B.
Figure 2.3. Non-specific proteins elute with NRF-1 in a gradient salt elution. Nuclear extracts
from MCF-7 cells were incubated with Gb-270multiBT (A) andGb-270(m10)multiBT (B) in an
immobilized-template assay. Bound proteins were isolated from nuclear extracts (unbound) and
rinsed with binding buffer (wash), prior to gradient elution using wash buffer containing 60-600
mM KCl, and a final elution (elute) with high heat. Isolated proteins were then subjected to
Western blot analysis with antibodies against ARID1A, Brg-1 and NRF-1.
30
increased separation of the elution profiles between NRF-1 and the non-specific proteins. In the
absence of magnesium (Figure 2.4a) and in the presence of 12.5 mM MgCl2 (Figure 2.4b), NRF1 was observed to elute between 300-500 mM KCl while the non-specific proteins were present
in fractions up to 300 mM KCl. Although the elution profiles were similar in both conditions, the
relative quantities of eluted proteins in the fractions differed. Within the elution range for NRF-1,
the majority eluted at the higher salt concentrations using the buffer without magnesium, and at
the lower salt concentrations in the presence of magnesium. Additionally, subsequent large scale
assays revealed increased retention of non-specific proteins which overlap with the elution of
NRF-1 in the presence of Mg2+, but is diminished in conditions without it (Figure 2.4c). In
contrast, protein binding in the presence of 12.5 mM MgCl2 and gradient salt elution without
magnesium resulted in considerable overlap between NRF-1 and Brg-1 (Figure 2.4d), suggesting
that binding and elution conditions without magnesium are optimal for minimizing non-specific
binding while maintaining NRF-1 and presumably complex formation.
Both protein-protein and protein-DNA interactions are known to be dependent on pH
which influences the binding energy resulting from changes in protonation state [124,125].
Therefore, the pH of the modified binding and elution buffer without magnesium was varied
between the physiological pH 7.0-8.0 to further minimize non-specific binding, assessed by Brg-1
retention, while maintaining NRF-1 binding at higher salt concentrations, which were examined
by Western blot analysis (Figure 2.4e,f). Non-specific and NRF-1 binding were shown to be
dependent on pH, as increases in pH from 7.0 to 8.0 were observed to decrease the DNA binding
activity of Brg-1 to the immobilized -268/-251 multimer, with reduced NRF-1 retention only
observed at pH 8.0. However, there was still substantial overlap in the elution of Brg-1 and NRF1 under these conditions. Based on these observations, the optimized immobilized template assay
31
B.
A.
- Mg2+
+ Mg2+
C.
D.
E.
F.
Figure 2.4. Optimization of elution conditions with Mg2+ and pH variation. Nuclear extracts
were incubated with Gb-270multiBT in an immobilized-template assay. Proteins were bound and
eluted with a gradient salt elution in Buffer Ze without magnesium (A), in Buffer Z with 12.5mM
magnesium (B), or bound in Buffer Z and eluted in Buffer Ze (D). Large scale immobilizedtemplate assays were performed with Buffer Ze or Buffer Z, and non-specific binding assessed by
Western blot analysis for Brg-1 (C). The pH of Buffer Ze was varied between pH 7.0-8.0 and
proteins were bound and eluted with a gradient salt elution (E) and (F). Isolated proteins were
subjected to Western blot analysis with antibodies against Brg-1, NRF-1 and TBP.
32
conditions to retain NRF-1 and likely the regulatory complex binding while minimizing nonspecific DNA binding consisted of reactions carried out at pH 7.5 using binding and elution
buffer without magnesium.
2.4.4 Isolated fractions containing NRF-1 are not consistent with complex formation
To verify the presence of the GABPβ -268/-251 regulatory complex in the fractions
containing NRF-1 isolated using the immobilized -268/-251 multimer with optimized conditions
(Figure 2.4e-f, pH 7.5), several EMSAs were performed. The -268/- 251 complex binding
sequence (Gb-270) and an oligonucleotide containing an NRF-1 binding site from the rat
cytochrome c promoter (RC4) [126] were utilized as probes. Initial EMSAs using these labeled
probes revealed some discordance between the NRF-1 containing fractions observed between
200-400 mM KCl using the RC4 probe, and those that produced a shift (C) with the Gb-270
probe which demonstrated complex formation between 100-200 mM KCl (Figure 2.5). This
discrepancy was not surprising, as it is possible that the GABPβ regulatory complex containing
NRF-1 dissociates from the immobilized -268/-251 multimer over multiple salt elution fractions.
This fractionated elution of the complex is supported by the different protein-protein and proteinDNA interactions that exist within this complex. It has been proposed that NRF-1 binds the -268/251 promoter sequence as a homodimer, with an additional unknown protein (X) making limited
contact with the upstream promoter and requiring interactions with NRF-1 for stability [65]. A
third unknown protein (Y) of the complex binds NRF-1 + X and has no contact with the
promoter. Additionally, it has been demonstrated that a partially assembled regulatory complex
containing only NRF-1 and protein X is able to shift Gb-270 in an EMSA, producing a faster
migrating complex than the entire NRF-1+X+Y regulatory complex [65]. Therefore, if protein X
was present in any fraction other than 100-200 mM KCl where a shift is already observed, then a
33
Figure 2.5. Fractions containing NRF-1 are not consistent with complex formation. Isolated
fractions from the immobilized-template assay with Gb-270multiBT and Buffer Ze (pH 7.5)
(Figure 2.4 e-f) were used in an EMSA with labeled oligonucleotide having a known NRF-1
binding site (RC4), and Gb-270 with the NRF-1 containing complex binding site. NRF-1 binding
(NRF-1), complex binding (C), non-specific binding (NS), and free probe (F) are indicated.
34
faster migrating complex containing NRF-1 would be observed. To test the possibility of a
fractionated complex during isolation, a reconstitution EMSA was performed whereby the NRF-1
containing 300 mM KCl fraction was added in equal volume to each of the gradient salt eluted
fractions, and the ability of each fraction to bind the GABPβ promoter was assessed using the Gb270 probe (Figure 2.6). Despite the addition of NRF-1 to the isolated fractions, there were no
additional shifts observed with the Gb-270 probe, as binding complexes (C) were again observed
in the 100-200 mM KCl fractions. The absence of additional shifts with NRF-1 supplemented
fractions suggests that either protein X and Y are only present in the 100-200 mM KCl fractions
with non-specific proteins where a shift is already observed, or the -268/-251 regulatory complex
was not isolated using the immobilized-template assay and is present in the unbound fraction,
with the shifts observed at the lower salt fractions the result of non-specific DNA binding
proteins.
2.4.5 Evaluation of binding specificity to -268/-251 by fractions isolated using the immobilizedtemplate assay
Despite the use of free competitor DNA to deplete non-specific DNA-binding proteins
and the optimization of binding and elution conditions, the presence of non-specific proteins
using immobilized-template assays is a frequent occurrence. To verify the specificity of binding
of the isolated complex to the GABPβ promoter at -268/-251, an EMSA was performed using a
mutated Gb-270 probe. The mutant probe (Gb-270 m6) contains conservative nucleotide
substitutions of residues -267/-265 which have been previously shown to disrupt formation of the
regulatory complex in an EMSA [65]. A shift of the Gb-270 m6 probe by the isolated fractions
but not with MCF7 whole cell lysate (WCL) positive control would indicate the enrichment of
non-specific DNA binding proteins in the isolated fractions. When the Gb-270 m6 probe was
35
Figure 2.6. NRF-1 supplemented fractions do not mediate additional complex formation.
Equal amounts of fractions isolated from the immobilized-template assay with Gb-270multiBT
and Buffer Ze (pH 7.5) and the NRF-1 containing 300 mM fraction were combined and used in
an EMSA with labeled RC4 and Gb-270 oligonucleotide. NRF-1 binding (NRF-1), complex
binding (C) and free probe (F) are indicated.
36
Figure 2.7. Complex binding is diminished upon binding site mutation. A Gb-270 m6
oligonucleotide with residues mutated to abolish complex binding was used in an EMSA with
fractions isolated from the immobilized-template assay with Gb-270multiBT and Buffer Ze (pH
7.5). Complex binding (C), non-specific binding (NS) and free probe (F) are indicated.
37
used in an EMSA with fractions isolated from the immobilized template assay (Figure 2.4 e-f,
pH7.5), diminished complex formation was observed with the MCF7 WCL positive control and
the isolated immobilized-template fractions (Figure 2.7).However, the absence of a shift in any of
the isolated fractions is not conclusive of binding specificity to Gb-270, especially as previous
Western blot analysis indicated the presence of the non-specific binding Brg-1 in fractions 200400 mM KCl (Figure 2.2F, pH 7.5). Therefore, these results collectively suggest that the binding
observed with Gb-270 is at least in part due to non-specific DNA binding proteins.
2.5 Discussion
GABP is involved in the transcriptional regulation of genes responsible for cell cycle
control, apoptosis, and differentiation [85]. With BRCA1 being one of its regulatory targets,
elucidation of the mechanisms regulating GABP expression could prove useful in the
understanding of human pathologies including sporadic breast cancers, as decreased GABPβ
promoter activation has been shown to result in low levels of BRCA1 in the tumorigenic SK-BR3 mammary cell line [65]. Our present study attempted to identify co-activators of the GABPβ
promoter that have been shown to form a regulatory complex with NRF-1 (NRF-1+X+Y)
approximately 270 bp upstream from the transcription start site (Gb-270). Our approach used an
immobilized-template affinity assay with the promoter-binding sequence of this complex.
Previous unpublished work suggested that the SWI/SNF subunit proteins ARID1A and Brg-1
formed part of this regulatory complex with NRF-1. A maltose-binding protein fused NRF-1
affinity column was initially used to isolate fractions containing NRF-1 binding proteins that
form a complex in the absence of a DNA scaffold. This was followed by an EMSA with Gb-270
to determine which fractions could bind the GABPβ promoter. These fractions were then
analyzed by mass spectrometry to identify possible binding proteins, which included ARID1A
and Brg-1. The SWI/SNF complex regulates gene expression through the modification of
38
chromatin structure using the energy derived from ATP hydrolysis [127]. As it has no intrinsic
ability to bind DNA, it requires transcription factors to recruit it to promoters [128]. This is
consistent with the proposed complex binding structure at the GABPβ promoter as NRF-1 has
been suggested to bind the GABPβ promoter while recruiting co-activators X and Y to the
promoter through protein-protein interactions. Evidence of its involvement in cancer comes from
the roles of several of its subunits, namely Brg-1, ARID1A and SMARCB1, which act as tumor
suppressors [129].
Despite these findings, our competition assay using the immobilized-template suggested
that these proteins were binding non-specifically (Figure 2.2). These results are supported by the
literature which maintains that unlike most ARID family proteins which preferentially bind ATrich sequences, ARID1A is able to bind DNA without sequence specificity [130,131].
Additionally, its affinity for non-specific DNA sequences is maintained up to at least 200 mM
KCl [130], which is consistent with our observations during the immobilized-template salt elution
(Figure 2.3). Furthermore, ARID1A is known to promote the formation of the SWI/SNF
chromatin remodeling complex containing the alternative catalytic cores Brg-1 or Brm [132], and
has been shown to interact directly with Brg-1 mediated by its helicase/SANT-associated (HSA)
domain [133]. It is also possible that Brg-1 is binding to the promoter through protein-protein
interactions with Sp1, which also appears to bind non-specifically. An interaction between Sp1
and Brg-1 has already been demonstrated at the SPARC promoter, where Sp1 was demonstrated
to recruit Brg-1 to the promoter [121]. Therefore, the non-specific binding of Sp1 could be
mediating additional binding of Brg-1 and ARID1A to the immobilized GABPβ promoter
through protein-protein interactions. Unfortunately, these non-specific DNA-binding proteins
were shown to be present in the fractions containing NRF-1 despite buffer optimization, with
39
some discrepancy between the fractions containing NRF-1 and those observed to complex at the
GABPβ promoter in an EMSA. This suggests that the regulatory complex proteins are
dissociating from the template over multiple isolated fractions, and that if the complex is present
in the 100-200 mM KCl fractions that are able to shift Gb-270, there are also contaminating nonspecific DNA-binding proteins present, which prevents mass spectroscopy analysis of the entire
fraction to identify the complex co-activators.
GABP and NRF-1 function as DNA-binding transcriptional regulators of mitochondrial
biogenesis, and frequently work in conjunction as regulators of nuclear-encoded mitochondrial
proteins required for oxidative phosphorylation [97,100]. As the protein-coding capacity of the
mitochondrial genome is limited, the expression of several nuclear-encoded genes is required
during mitochondrial biogenesis [97]. Therefore, coordinated expression of the nuclear and
mitochondrial genomes during mitochondrial biogenesis requires the coordinated induction and
regulatory activity of several proteins including GABP and NRF-1. As a complex containing
NRF-1 has been shown to regulate the GABPβ promoter, it is likely that the unidentified coactivators that complexes with NRF-1 are also involved in the regulation of mitochondrial
biogenesis.
The peroxisome proliferator-activated receptor gamma co-activator (PGC)-1 family
members (PGC-1α, PGC-1β and PGC-1 related co-activator (PRC)) are nuclear transcriptional
co-activators that serve as key regulators of the transcriptional network controlling mitochondrial
biogenesis and function [98]. This family of co-activators interacts with a number of nuclear
DNA-binding transcription factors and hormone receptors including NRF-1, GABP, ERRα, and
PPARα/γ among others to integrate nuclear gene expression with mitochondrial biogenesis
[97,98]. Through direct binding to NRF-1 in mouse myoblasts, PGC-1α has been shown to
40
transcriptionally activate several NRF-1 target genes, with a dominant negative NRF-1 lacking
the trans-activation domain able to abrogate this effect [104]. PGC-1β requires an interaction
with both NRF-1 and ERRα to activate several NRF-1 target genes, where both transcription
factors are required for maximal promoter activation [134]. PRC has been show to interact with
NRF-1 to activate the cytochrome c promoter in vitro [135]. As a direct regulatory interaction
between NRF-1 and all members of the PGC-1 family of co-activators has already been shown to
exist, it is possible that one of these co-activators forms part of the regulatory complex with NRF1 at the GABPβ promoter. Additionally, decreased expression of PGC-1 co-activators have been
observed in human breast [136,137], colon [138] and ovarian [139] tumor samples, and given
their role in mitochondrial biogenesis, disruption of PGC-1 co-activator function has been shown
to result in mitochondrial dysfunction in a number of diseases [140,141]. Interestingly, cancer
cells exhibit an altered metabolic program known as the Warburg effect, whereby there is a
preference to perform glycolysis over oxidative phosphorylation despite a sufficient oxygen
supply to support mitochondrial metabolism. Although it has yet to be demonstrated, disrupted
PGC-1 co-activator expression could contribute to the low BRCA1 levels characteristic of
sporadic breast cancers through its aberrant interaction with NRF-1 at the GABPβ promoter,
while also contributing to the altered metabolism observed in breast tumor cells. Therefore,
further experimentation is needed to support a role for PGC-1 co-activators in this regulatory
network.
Identifying DNA-binding transcription factors and protein complexes based on their
affinity for DNA using an immobilized-template is a widely used approach, and has been
successfully used in the isolation of transcription factors including Sp1 and activator protein 1
(AP-1) [142], and protein complexes including the NFAT transcriptional complex [143]. The
41
isolation of Sp1 and AP-1 involved the use of competitor DNA during two successive affinity
purification steps with the oligonucleotide containing the DNA-binding sequence bound to
CNBr-activated Sepharose. This assay utilized buffer containing 100 mM KCl, 12.5 mM MgCl2,
and a pH 7.8, with a gradient salt elution of bound proteins. Although Sp1 and AP-1 were
purified to near homogeneity, there were still non-specific binding proteins present. Further
purification steps including the use of additional DNA-Sepharose resins with oligonucleotides
having different flanking sequences would be necessary to isolate a homogeneous solution of the
desired protein. In contrast to the purification of already identified transcription factors, the
isolation of novel co-activators in a protein complex can be exemplified by Yang and Chow
[143]. The identification of PARP-1 as an NFAT transcriptional co-activator important in
immune response was achieved using an immobilized biotin-labeled PPARγ2 proximal NFATbinding oligonucleotide in the presence of poly (dI-dC). A competition assay similar to ours was
also performed using free oligonucleotide to determine the specificity of isolated proteins. In both
assays the bound proteins were immediately eluted and resolved using SDS-PAGE, and stained.
Proteins binding specifically to the oligonucleotide demonstrated reduced binding in the presence
of competitors, and following gel excision and digestion with trypsin, were identified using
MALDI-TOF (matrix-assisted laser desorption/ionization – time of flight) mass spectrometry.
This was necessary as there was substantial non-specific binding observed in their assay eluate.
Therefore, although it appears our strategy using an immobilized-template assay with optimized
conditions is valid, it remains incomplete. Our aim to analyze the entire isolated fraction was
plagued with the presence of non-specific DNA-binding proteins including ARID1A and Brg-1
that eluted from the immobilized-template in the same fraction as NRF-1 and presumably the coactivators of the regulatory complex. However, additional experimentation that would include a
42
comparison between proteins immediately eluted from the immobilized-template in the presence
and absence of GABPβ competitors would reveal which proteins are binding specifically. These
proteins could be resolved using SDS-PAGE, excised and identified by mass spectroscopy, or the
entire fractions obtained with and without competitors could be analyzed with a subtractive
assessment of the fraction contents. The latter would ensure that all bound proteins would be
accounted for and would circumvent the reliance on protein visualization and gel excision
techniques.
2.6 Conclusions
The isolation and identification of co-activators at the GABPβ promoter that form a regulatory
complex with NRF-1 was attempted using an immobilized-template assay. While the assay
isolated proteins including NRF-1 that could bind the promoter, there was substantial nonspecific binding observed that could not be abolished through optimization of binding and elution
conditions, which prevented identification of the isolated proteins using mass spectroscopy.
Therefore, further experimentation is needed to distinguish between the non-specific DNAbinding proteins and the co-activators that complex with NRF-1 at the GABPβ promoter.
2.7 Acknowledgements
We would like to thank Crista Thompson, Rachael Klinoski, Valerie Kelly-Turner, and Sherri
Nicol for their excellent technical assistance. This work was funded by a grant from the Canadian
Breast Cancer Foundation – Ontario Region.
43
CHAPTER 3: The role of NRF-1 in BRCA1-mediated differentiation and mammary gland
tumourigenesis
3.1 Abstract
Background: BRCA1 is a tumor suppressor that contributes to the maintenance of genome
integrity, with a regulatory role in luminal progenitor differentiation. Decreased expression of
BRCA1 is characteristic of basal-like tumors, which arise from undifferentiated luminal
progenitors. NRF-1 and GABPβ have been shown to function in a regulatory network with
BRCA1, while also regulating the expression of genes required for mitochondrial metabolism.
While both breast differentiation and mitochondrial metabolism are frequently disrupted in breast
cancer, it remains unclear whether there is a direct role for NRF-1 in BRCA1-mediated
differentiation.
Results: In the present study, we utilized a doxycycline-inducible lentiviral system to generate
NRF-1 knockdown cell lines to examine its effect on differentiation mediated by BRCA1. When
these cells differentiate, there is significant infilling of the luminal space and altered cell
polarization in the observed cell structures. In monolayer cells, NRF-1 knockdown does not result
in decreased BRCA1 expression. However, evaluation of the mitochondrial metabolism of these
NRF-1 depleted cells revealed notable mitochondrial dysfunction. Furthermore, there was
decreased expression of several NRF-1 targets genes involved in mitochondrial metabolism and
mitochondrial-induced apoptosis, which are likely required for maintenance of a hollow luminal
space.
Conclusion: Overall, our results reveal a potential role for NRF-1 in mediating apoptosis during
mammary ductal morphogenesis through the maintenance of normal mitochondrial function, and
44
although its activity has been shown to moderate BRCA1 expression, the involvement of NRF-1
in BRCA1-mediated differentiation remains undetermined.
3.2 Background
The mammary gland structure is derived primarily from epithelial progenitor cells of
basal or luminal lineage which gives rise, respectively, to cells forming the exterior or lining the
interior of hollow mammary ducts [144]. The basal lineage generates a cell layer which consists
primarily of differentiated myoepithelial cells [145], with several mammary stem cells also
embedded within this layer [21,144]. Cells of luminal origin include ductal and alveolar cells, the
latter of which arises during pregnancy and serves as the milk-producing unit during lactation
[144]. The tumor-suppressor gene BRCA1 (breast cancer 1, early onset) has been proposed to
regulate luminal progenitor cell differentiation, with reduced expression shown to result in an
accumulation of undifferentiated stem and progenitor cells which are suggested to serve as prime
targets for tumourigenesis [36,64]. Furthermore, knockdown of BRCA1 impairs the formation of
mammary acini in an in vitro 3D culture which mimics in vivo mammary ductal morphogenesis,
which supports a role for BRCA1 in mammary differentiation [63]. Breast tumors associated with
a significant reduction in BRCA1 expression characteristically have a basal-like phenotype, and
have been shown to originate from undifferentiated luminal epithelial progenitors [35,146].
Therefore, determining the basis for the decreased BRCA1 expression in breast tumors could
provide insight into the etiology of these undifferentiated basal-like tumors.
The GA-binding protein (GABP) is a well characterized transcriptional regulator of
BRCA1 expression [45], and exists as an obligate multimer consisting of a pair of GABPα/β
heterodimers where GABPα seves as the DNA-binding subunit and GABPβ mediates
transactivation
[85,86].
Tetramerization
of
45
two
GABPα/β
heterodimers
forms
the
transcriptionally active GABP, which has been demonstrated as being a crucial regulator of
several genes involved in the maintenance and differentiation of hematopoietic stem and
progenitor cells [92]. This role for GABP can likely be extended to the mammary gland through
its regulation of BRCA1, which has been determined to be crucial for mammary stem and
progenitor cell differentiation [64]. Recently, BRCA1 expression has also been shown to be
dependent on Nuclear Respiratory Factor 1 (NRF-1), which regulates the GABPβ promoter [65].
Interestingly, GABP and NRF-1 coordinately regulate the expression of several nuclear encoded
genes that are essential components of mitochondrial metabolism [97]. Breast tumors frequently
demonstrate a shift in metabolism [147] where they primarily metabolize glucose by aerobic
glycolysis, rather than the mitochondrial oxidative phosphorylation performed by normal cells; a
phenomenon known as the Warburg effect [109]. Therefore, down-regulation of BRCA1 through
reduced NRF-1/GABPβ expression could result in an accumulation of undifferentiated luminal
progenitor cells, in addition to a shift towards the classical metabolic phenotype of cancer cells,
providing a link between differentiation and cell metabolism.
In this study, we set out to investigate the regulatory role of NRF-1 during BRCA1mediated differentiation of pluripotent mammary cell lines, and to evaluate the importance of
mitochondrial metabolism in this process. We show that reduction of NRF-1 expression by an
inducible lentiviral shRNA causes a failure of mammary acinar formation in 3D culture, but in
2D culture does not result in a sustained decrease in expression of GABPβ/BRCA1. We also
describe the mitochondrial dysfunction resulting from NRF-1 depletion, which accompanies
decreased expression of several nuclear encoded mitochondrial genes potentially involved in
hollow lumen formation and maintenance through the induction of caspase-dependent apoptosis.
The data presented here suggest that NRF-1 plays a role in mammary ductal morphogenesis
46
potentially through its regulation of mitochondrial metabolism and caspase-dependent cell death,
with its role in BRCA-1 mediated differentiation requiring further investigation.
3.3 Methods
3.3.1 Cells and cell culture
The human non-tumorigenic mammary epithelial cell line MCF-10A was obtained from the
ATCC (Manassas, VA, USA), while the 184hTERT cell line [148] was a gift from Dr. Calvin
Roskelley. MCF-10A cells were maintained in DMEM-F12 with L-glutamine (HyClone)
supplemented with 5% horse serum (Invirogen, Burlington, Canada), 20 ng/ml epidermal growth
factor (Invitrogen), 10 μg/ml insulin (Sigma), 0.5 μg/ml hydrocortisone (Sigma), 100 ng/ml
cholera toxin (Sigma), 100 units/ml penicillin and 100 μg/ml streptomycin. 184hTERT cells were
maintained in MEBM media (Clonetics) supplemented with SingleQuot kit (Clonetics), 400
μg/ml G418 (BioShop, Burlington, Canada), 1 μg/ml transferrin (BD, Mississauga, Canada), and
1.25 μg/ml isoproterenol (Sigma). Both cell lines were maintained at 37oC with 5% CO2.
3.3.2 Lentiviral production and transduction
HEK 293T cells obtained from Dr. David LeBrun were plated for transfection at 5.5 x 106
cells/100mm plate in DMEM supplemented with 10% fetal bovine serum, 100 units/ml penicillin,
and 100 g/ml streptomycin, and humidified at 37oC with 5% CO2 for 24 hours. HEK293T cells
were transfected with the following plasmids using the Fugene 6 transfection reagent
(DNA:Fugene, 1:4) (Promega): packaging vector psPAX2 (11.3 μg/100mm plate), envelope
vector pMD2.G (5.7 μg/100mm plate) and either non-silencing pTRIPZ-shRNAmir control
(RHS4743) or NRF-1 targeted silencing pTRIPZ-shRNAmir (V2THS_20255) (17 μg/100mm
plate) from Open Biosystems (Thermo Fisher Scientific). Lentiviral supernatants were collected
after 48 and 72 hours post-transfection and immediately filtered with a 0.45 μm PVDF filter and
47
frozen at -80oC, producing viral titers of 105-106 transducing units/ml. 184hTERT and MCF-10A
cells were plated at 5 x 104 cells/ml in 100mm plates and incubated at 37oC with 5% CO2 for 24
hours. Cell media was removed from the plates and 5 ml serum-free media supplemented with 10
μg/ml polybrene (H9268, Sigma) was added. Cells were transduced with the lentiviral
supernatant shRNA-control at a multiplicity of infection (MOI) of 1 and 2 (named ctrl-m1 or ctrlm2) or shRNA-NRF-1 at a multiplicity of infection (MOI) of 1, 2 and 6 (named shNRF1-m1, m2, -m6) added directly to the serum-free media. Following 4 hours incubation at 37oC with 5%
CO2, media was removed and replaced with 10 ml regular culture media. Cells were incubated for
24 hours at 37oC with 5% CO2 and media was replaced with regular culture media containing 0.5
μg/ml puromycin to select for cells that stably express the shRNA plasmid. The shRNA was
induced by treating cells with 1 μg/ml doxycycline with daily replenishment directly to cells in
culture media.
3.3.3 Acini
The three-dimensional culture of 184hTERT and MCF-10A cells on Matrigel was performed as
previously described [149], with the following modifications: Matrigel was spread onto precooled 12-well plates prior to incubation at 37oC with 5% CO2 in a cell culture incubator to
solidify the basement membrane, after which a total of 3500 cells/well were plated in assay media
(DMEM-F12 containing 2% horse serum, 100 ng/ml cholera toxin, 0.5 μg/ml hydrocortisone, 10
μg/ml human insulin, 100 units/ml penicillin, and 100 g/ml streptomycin, and 5 ng/ml epidermal
growth factor) containing 2% Matrigel overlay. For induction of transduced lentiviral cells with
doxycycline, the assay media overlay was supplemented with 1 μg/ml doxycycline with daily
replenishment.
48
3.3.4 Preparation of whole cell lysates
For NRF-1 knockdown time course, cells were plated in a 6-well plate at 8.75 x 104 cells/well or
in 100 mm plates at 3.5 x 105 cells/plate in regular culture media. Cells were incubated for 24
hours at 37oC with 5% CO2 after which cells were put into puromycin selection and appropriate
wells/plates were treated with doxycycline with daily replenishment. Seventy-two hours
following doxycycline treatment, cells were lysed using 1 x SDS-PAGE loading buffer (2.5%
sodium dodecyl sulfate, 25mM Tris-HCl pH 6.8, 100mM DTT, 10% glycerol) supplemented with
protease and phosphatase inhibitors (1 mM EDTA, 1 mM PMSF, 1 μg/ml each of aprotinin,
leupeptin and pepstatin, 1 mM sodium orthovanadate, 1 mM NaF) and frozen at -20oC.
3.3.5 Western blot analysis
Whole cell lysates were resolved on a SDS-polyacrylamide gel, transferred to PVDF membrane,
and probed with the appropriate antibody. Primary antibodies included: anti-NRF-1 (M01, 1:250,
Abnova), anti-TBP (1TBP18, 1:2000, Abcam), anti-BRCA1 (0P92, 1:500, Calbiochem, San
Diego, CA, USA), anti-GABPα (H-180, 1:500, Santa Cruz Biotechnology) and anti-GABPβ (H265, 10x, 1:5000, Santa Cruz Biotechnology). Secondary antibody detection was performed using
SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific/Fisher, Nepean, Canada).
3.3.6 Quantitative real-time PCR (qRT-PCR)
RNA was isolated using the Genelute Mammalian Total RNA Miniprep Kit (Sigma). Quantitative
real-time PCR (qRT-PCR) reactions were carried out using specific TaqMan gene expression
probes for human NRF-1 (Hs00192316_m1, FAM), BRCA1 (Hs01556193_m1, FAM), GABPα
(Hs01022016_m1, FAM), GABPβ (Hs00242573_m1, FAM), Tfam (Hs00273372_s1, FAM), and
cytochrome c (Hs01588974_g1, FAM) (all from Life Technologies Incorporated). Human
hypoxanthine phosphoribosyltransferase 1 HPRT-1 (Hs99999909_m1, VIC, Life Technologies
49
Incorporated) was used as a normalization control. Reactions were performed using the
SuperScript® III Platinum® One-Step Quantitative RT-PCR System (Invitrogen) with 25 ng RNA
per reaction and TaqMan probes specified above. Reverse transcription of RNA to cDNA and
subsequent amplification was done sequentially in one tube. The PCR protocol consisted of 1
cycle for cDNA conversion and enzymatic activation (15 min at 50oC and 2 min at 95oC)
followed by 40 cycles (15 sec at 95oC, 30 sec at 60oC). Gene expression was expressed relative to
the wild-type control using the delta-delta Ct method described by PE Applied Biosystems
(Perkin Elmer, Forster City, CA, USA).
3.3.7 Dual luciferase assay
Approximately 48 hours prior to transfection, uninduced lentiviral cells (184hTERT and MCF10A shNRF1-m1) were plated in 12-well plates at 3.5 x 104 cells/well. Cells were incubated for
24 hours at 37oC with 5% CO2 after which 72 hr induced shNRF1-m1 cells were treated with 1
μg/ml doxycycline; 24 hr induced shNRF1-m1 and uninduced received no treatment. Prior to
transfection, media was replaced for all cells with regular culture media without doxycycline.
Cells were transfected in triplicate using a total of 250 ng DNA per well with 0.75 μl/well
FuGENE 6 (Roche, Laval, Canada) according to the manufacturer’s protocol. In each well, 25 ng
cytomegalovirus (CMV)-luc internal control vector (Promega) and 225 ng of the appropriate
Renilla luciferase GABPβ reporter vector (pRL) was added. Approximately 6 hours posttransfection, 1 μg/ml doxycycline was added directly to the cell media for 72 hr shNRF1 induced
cells only. Approximately 24 hours post-transfection, 24 hr and 72 hr shNRF1 induced cells
received 1 μg/ml doxycycline treatment. Approximately 48 hours post-transfection, all cells were
washed with phosphate buffered saline (PBS), lysed in 150 μl passive lysis buffer (Promega), and
50
assayed with the Dual-luciferase® Reporter Assay System according to the manufacturer’s
protocol (Promega) using an EG&G Berthold microplate luminometer.
3.3.8 Mitochondrial function assay
184hTERT and MCF-10A wild-type and shNRF1-m1 cells (uninduced and 72 hr doxycycline
induced) were plated in duplicate using regular culture media in 6-well plates at 3.5 x 104
cells/well and allowed to incubated for 24 hours at 37oC with 5% CO2. Media was removed and
replaced with reduced culture media for 184hTERT: RPMI 1640 without glucose [R6504,
Sigma], supplemented with 2.0 g/L sodium bicarbonate (Sigma), 10% fetal bovine serum, 100
units/ml penicillin, and 100 g/ml streptomycin; and MCF-10A: DMEM without glucose or Lglutamine [D5030, Sigma], supplemented with 0.584 g/L of L-glutamine, 3.7 g/L sodium
bicarbonate, 5% horse serum (Invitrogen, Burlington, Canada), 100 ng/ml cholera toxin (Sigma),
100 units/ml penicillin and 100 μg/ml streptomycin) containing 10 mM glucose or galactose.
Cells were maintained in appropriate treatment (1 μg/ml doxycycline or DIUF H2O vehicle), with
reduced culture media being replaced daily. Cells were lifted using 1x trypsin-EDTA and counted
using a hemocytometer following 1, 2, 3 and 5 days of growth in reduced culture media
containing glucose or galactose.
3.4 Results
3.4.1 Mammary epithelial cells with decreased NRF-1 expression demonstrate abnormal aciniar
morphogenesis
NRF-1 has been previously shown to regulate the GABPβ promoter, and its aberrant
activation results in decreased BRCA1 expression [65]. As luminal cell differentiation is
dependent on BRCA1 expression [64], we evaluated the role of NRF-1 in mammary epithelial
cell differentiation. Inducible NRF-1 knockdown was achieved through transduction of human
51
non-tumorigenic mammary epithelial cells (184hTERT and MCF-10A) with increasing amounts
of shRNAmir lentiviral vectors (MOI of 1, 2, and/or 6) encoding a non-silencing control or a
doxycycline (Dox) inducible shRNA against NRF-1. Stable inducible 184hTERT and MCF-10A
cell lines are referred to as ctrl-m1 and ctrl-m2 (non-silencing control MOI 1 and MOI 2), and
shNRF1-m1, -m2, -m6 (shRNA against NRF-1 MOI 1, 2, 6). In shNRF1-m1, -m2, and m-6 for
both cell lines, a gradient knockdown of NRF-1 was generally observed both with Dox induction
and increasing MOI at the protein (Figure 3.1 a) and mRNA level (Figure 3.2). Induced
184hTERT shNRF1-m1 showed a downregulation of 29% in NRF-1 mRNA, followed by
shNRF1-m6 (45%), and shNRF1-m2 (70%) (Figure 3.2 a), while MCF-10A shNRF1-m1 showed
a downregulation of 29% in NRF-1 mRNA, followed by shNRF1-m6 (30%), and shNRF1-m2
(59%) (Figure 3.2 b). The 184hTERT and MCF-10A ctrl-m1 and ctrl–m2 produced no substantial
effect on NRF-1 protein (Figure 3.1 b) and mRNA expression (Figure 3.2), with the apparent
decrease in NRF-1 protein for MCF-10A ctrl-m2 attributed to transfer issues during the Western
blot analysis. However, it was obvious from the qRT-PCR analysis that the uninduced (UT)
shNRF1-m1,-m2,-m6 in both cell lines resulted in a decrease in NRF-1 expression varying
between 15-30% knockdown in the absence of Dox induction compared to non-manipulated wildtype cells, indicating that the shRNA is ‘leaky’.
The morphological development of the mammary gland can be modeled in vitro through
the induced differentiation of 3D mammary acini by growth on Matrigel, where a single epithelial
cell layer polarizes apico-basally in growth-arrested spheroids with hollow lumens [149,150].
Depletion of BRCA1 during cellular differentiation has been shown to result in impaired acinar
52
A.
B.
Figure 3.1. Inducible lentiviral mediated knockdown of NRF-1 protein. 184hTERT and
MCF10A cells were transduced with lentivirus expressing NRF-1 shRNA (shNRF1) (A) or nonsilencing shRNA control (ctrl) (B) to generate stable lines. A multiplicity of infection (m) of 1, 2,
and 6 was used for shNRF1 and m1 and m2 for ctrl lines. Cells were grown in the absence (UT)
or in the presence of 1ug/mL dox for 24-72 hours with daily replenishment. Whole cell lysates
were prepared and normalized, and Western blot analysis was performed using an antibody
against NRF-1 for the indicated cell lines and MOIs; an arrow indicates NRF-1 protein. TBP was
used as a loading control to normalize for overall protein concentration. Western blots for
shNRF1 are representative of three experiments, and ctrl is representative of one experiment.
53
1.50
Relative Expression
A.
UT
72
1.25
1.00
0.75
0.50
0.25
6
sh
N
R
F1
-m
2
F1
sh
N
R
F1
sh
N
R
ct
B.
-m
1
-m
2
rlm
1
rlm
ct
W
T
0.00
1.75
UT
72
Relative Expression
1.50
1.25
1.00
0.75
0.50
0.25
m
6
sh
N
R
F1
-
m
2
sh
N
R
F1
R
N
sh
F1
-
m
1
m
2
ct
rl-
m
1
ct
rl-
W
T
0.00
Figure 3.2. Inducible lentiviral mediated knockdown of NRF-1 mRNA. 184hTERT shNRF1m1,-m2,-m6 and ctrl-m1,-m2 (A), and MCF-10A shNRF1-m1,-m2,-m6 and ctrl-m1,-m2 (B) were
grown in the absence (UT) or in the presence of 1ug/mL dox for 72 hours with daily
replenishment. RNA was prepared and NRF-1 expression was analyzed by qRT-PCR for
184hTERT and MCF10A cell lines. Values were normalized relative to an HPRT internal control
and fold changes are expressed relative to a non-manipulated wild type (WT) control (relative
expression set to 1.0). qRT-PCR analysis is representative of two experiments and values
represent averages of triplicate samples +/- SD.
54
formation, where irregular shaped multilobular structures with filled lumens are observed [63].
To determine the effect of NRF-1 depletion on acinar formation, non-manipulated wild type
(WT), induced ctrl-m1, uninduced shNRF1-m1, and induced shNRF1-m1,-m2,-m6 cells were
plated on Matrigel, and their differentiation monitored over 15 days for 184hTERT (Figure 3.3)
and MCF-10A (Figure 3.4) cell lines. After 3 days of growth, 184hTERT and MCF-10A WT,
induced ctrl-m1 and uninduced shNRF1-m1 acini were spherical in shape and the outer cells had
begun to polarize, with few irregular structures observed. In contrast, induced shNRF1-m1,-m2,m6 acini appeared as globular clusters having outward projections, which has been suggested to
reflect a migratory behavior or potentially to mediate cell communication [63]. After 12 days,
WT and induced ctrl-m1 had nearly formed mature acini with a single layer of polarized
columnar shaped epithelial cells surrounding a hollow lumen. Uninduced shNRF1-m1 cells also
formed a hollow lumen but additionally had an irregularly polarized epithelial layer with cells
often protruding from the polarized layer boundary (Figure 3.3, 3.4). The slight morphological
difference between the WT/induced ctrl-m1 and uninduced shNRF1-m1 acini is attributed to the
‘leaky’ shRNA expression against NRF-1 that was observed with the qRT-PCR analysis (Figure
3.2). The induced shNRF1-m1,-m2,-m6 acini also appeared as globular clusters with filled
lumens and no clear distinction between the outer and inner luminal cells, indicating a lack of cell
polarization. This infilling and globular morphology is similar to what is observed in acini with
depleted BRCA1 [63], and with the activated oncogene ErbB2 [151].
To quantify the infilling observed in the 184hTERT and MCF-10A shNRF1-m1,-m2,-m6
acini, the nuclei of acini at day 15 were stained using Hoescht nuclear stain (Figure 3.5a) and the
proportion of acini with filled lumens was determined (Figure 3.5b). With Dox induction, a large
fraction of acini (~75%) exhibited filled lumens with this proportion remaining relatively constant
55
Uninduced
184hTERT WT
ctrl-m1
shNRF1-m1
shNRF1-m1
shNRF1-m2
shNRF1-m6
Figure 3.3. NRF-1 knockdown impairs mammary cell differentiation in 184hTERTs. Cells were induced to differentiate and form acini with
growth on Matrigel for 15 days. shNRF1-m1,-m2,-m6 were maintained in 1ug/mL dox with daily replenishment and compared with uninduced
shNRF1-m1, induced ctrl-m1 and non-manipulated wild type (WT) 184hTERT cells. Bright field images of the cells at Day 3 and 12 with 10x/40x
magnifications are shown (scale bars 50 μm and 100 μm). Differentiaton assay pictures are representative of two experiments.
56
Uninduced
MCF-10A WT
ctrl-m1
shNRF1-m1
shNRF1-m1
shNRF1-m2
shNRF1-m6
Figure 3.4. NRF-1 knockdown impairs mammary cell differentiation in MCF-10As. Cells were induced to differentiate and form acini with
growth on Matrigel for 15 days. shNRF1-m1,-m2,-m6 were maintained in 1ug/mL dox with daily replenishment and compared with uninduced
shNRF1-m1, induced ctrl-m1 and non-manipulated wild type (WT) MCF-10A cells. Bright field images of the cells at Day 3 and 12 with 10x/40x
magnifications are shown (scale bars 50 μm and 100 μm). Differentiaton assay pictures are representative of two experiments.
57
A.
B.
MCF-10A
184hTERT
100
100
75
WT
UT
DOX
50
% Infilling
% Infilling
75
WT
UT
DOX
50
25
25
0
0
1
2
1
6
2
6
MOI
MOI
Figure 3.5. NRF-1 deficient cells exhibit impaired lumen formation. The nuclei of Day 15
differentiated 184hTERT and MCF-10A WT and shNRF1-m1,-m2,-m6, which were either
untreated (UT) or maintained in 1ug/mL doxycycline, were visualized with Hoescht nuclear stain.
Representative images of stained 184hTERT cells were imaged at 40x magnification using a
DAPI filter (scale bar 50 μm) (A). Infilling in both cell lines was quantified by counting hollow
lumen formation in 200 acini, and compared with infilling present in non-manipulated wild-type
(WT) acini exposed to doxycycline during differentiation (B). Differentiation assay was repeated
twice and values represent average results of duplicate samples for each MOI +/- SD.
58
among the MOIs, while without Dox induction there was an increasing percentage of infilling
associated with increasing MOI (25-60%). This increased infilling trend in the uninduced
shNRF1 acini is again attributed to the increasing amount of ‘leaky’ shRNA present in these cells.
Although some luminal infilling was observed in WT acini, the incidence was generally less than
25% for both cells lines (Figure 3.5b). Interestingly, hollow lumen formation is still observed in a
small proportion of acini with NRF-1 depletion, although likely a result of cell variability in the
expression of the NRF-1 targeted shRNA. Overall, these observations suggest that NRF-1 is
essential for the normal aciniar morphogenesis of mammary epithelial cells.
3.4.2 NRF-1 depletion does not result in a long-term depression of GABPβ/BRCA1 expression in
monolayer cells
Because transient NRF-1 knockdown has been shown to attenuate GABPβ and BRCA1
gene expression [65], the infilling phenotype associated with NRF-1 depletion may be the result
of decreased BRCA1 expression, which has already been shown to result in infilled, multilobular
acinar structures [63]. To determine whether stable NRF-1 knockdown results in a prolonged
decrease in the expression of GABPα, GABPβ and BRCA1, we assessed their expression through
Western blot and qRT-PCR analysis in monolayer 184hTERT (Figure 3.6) and MCF-10A (Figure
3.7) cells over 72 hours of induced NRF-1 knockdown using samples analyzed in Figures 3.1 and
3.2. Although there was an initial decrease in GABPα and β expression at both the mRNA and
protein level following 24 hours of NRF-1 knockdown in 184hTERT shNRF1-m1,-m2,-m6 cells,
the expression generally recovered to uninduced levels by 72 hours (Figure 3.6 a,c-d). As NRF-1
is not a known regulator of GABPα, it is likely that the initial decrease in expression is the result
of low GABPβ expression which it requires for stabilization of the α/β heterodimer [65,86], and
the subsequent increase is a result of a positive feedback loop that serves to maintain steady-state
59
shNRF1-m6
MOI 6
B.
- 63 kDa
GABPα
GABPβ
- 63 kDa
TBP
- 48 kDa
Relative Expression
- 245 kDa
BRCA1
UT
24
48
72
2.0
48
72
UT
24
48
72
UT
24
48
72
shNRF1-m2
MOI 2
UT
24
shNRF1-m1
MOI 1
A.
1.5
1.0
0.5
0.0
shNRF1-m1
Relative Expression
1.0
0.0
shNRF1-m2
UT
24
48
72
3.0
0.5
shNRF1-m1
shNRF1-m6
D.
UT
24
48
72
1.5
Relative Expression
C.
shNRF1-m2
2.5
2.0
1.5
1.0
0.5
0.0
shNRF1-m6
shNRF1-m1
shNRF1-m2
shNRF1-m6
Figure 3.6. NRF-1 depletion does not result in a prolonged decrease of GABPα, GABPβ and BRCA1 expression in monolayer 184hTERT
cells. Western blot (A) and qRT-PCR analysis of BRCA1 (B), GABPα (C), and GABPβ (D) expression levels in 184hTERT shRNA-m1,-m2,-m6
over 72 hours of 1ug/mL doxycycline induction to decrease NRF-1 expression. Whole cell lysates were collected and normalized for overall
protein concentration with TBP loading control, and RNA was prepared from the indicated treatment conditions with an HPRT internal control.
qRT-PCR fold changes are expressed relative to a non-manipulated wild type (WT) control (relative expression set to 1.0) and values represent
averages of one experiment with triplicate samples +/-SD. Western Blot figures are representative of two experiments.
60
B.
BRCA1
- 245 kDa
GABPα
- 63 kDa
GABPβ
- 63 kDa
P
- 48 kDa
UT
24
48
72
C.
Relative Expression
1.25
1.00
0.75
0.50
0.25
1.50
1.25
1.00
0.75
0.50
0.25
0.00
shNRF1-m1
shNRF1-m2
shNRF1-m6
UT
24
48
72
1.50
D.
Relative Expression
TBP
UT
24
48
72
1.75
Relative Expression
72
48
shNRF1-m6
MOI 6
UT
24
72
U
T
24
shNRF1-m2
MOI 2
72
48
U
T
24
shNRF1-m1
MOI 1
48
A.
1.25
1.00
0.75
0.50
0.25
0.00
0.00
shNRF1-m1
shNRF1-m2
shNRF1-m6
shNRF1-m1
shNRF1-m2
shNRF1-m6
Figure 3.7. NRF-1 depletion does not result in a prolonged decrease of GABPα, GABPβ and BRCA1 expression in monolayer MCF-10A
cells. Western blot (A) and qRT-PCR analysis of BRCA1 (B), GABPα (C), and GABPβ (D) expression levels in MCF-10A shRNA-m1,-m2,-m6
over 72 hours of 1ug/mL doxycycline induction to decrease NRF-1 expression. Whole cell lysates were collected and normalized for overall
protein concentration with TBP loading control; non-phosphorylated GABPβ is indicated with an arrow and phosphorylated GABPβ is indicated
by (P). RNA was prepared from the indicated treatment conditions with an HPRT internal control. qRT-PCR fold changes are expressed relative to
a non-manipulated wild type (WT) control (relative expression set to 1.0) and values represent averages of one experiment with triplicate samples
+/-SD. Western Blot figures are representative of two experiments.
61
GABPα expression [94]. Expression of BRCA1 followed a trend similar to the expression of
GABPα and β (Figure 3.6 a,b), consistent with their critical role in BRCA1 promoter regulation.
Induced MCF10A shNRF1-m1,-2,-6 generally behaved similar to 184hTERT cells, although with
a smaller relative change in expression (Figure 3.7 a-d). A comparison between endogenous
NRF-1, GABPα/β and BRCA1 RNA levels in wild-type 184hTERT and MCF-10A cells revealed
that the relative expression was generally consistent between the cell lines, with the exception of
GABPβ mRNA in MCF-10A cells which was nearly two fold less than in 184hTERTs (Figure
3.8). This low GABPβ expression in MCF-10As could be contributing to the muted changes in
relative expression as compared with 184hTERT cells. Although the resultant increase in
expression of these regulatory network members in both cell lines plated in monolayer is not the
result of NRF-1 expression, which remained low at 72 hours (Figure 3.1 and 3.2), we cannot
extend these conclusions to explain the infilling defect observed following NRF-1 depletion in
3D acini. This is because while gene expression generally remains constant in monolayer cells, it
varies during the differentiation of 3D acini; a trend we have already observed with NRF-1,
GABP and BRCA1 (unpublished).
To investigate the molecular basis for the increased GABPβ expression despite persistent
NRF-1 knockdown in both cell lines, the activity of the GABPβ proximal promoter from -1023 to
+194 was examined using a series of 5’ promoter deletion constructs prepared using luciferase
reporter plasmids [65]. As NRF-1 has been shown to regulate the GABPβ promoter activity
through a site between -268/-251 [65], depletion of NRF-1 would be expected to result in
decreased activity of the entire promoter. Indeed, the promoter activity of MCF-10A shNRF1-m1
was decreased by 3-6 fold in comparison to WT cells (data not shown). Additionally, as expected
there was approximately a two-fold decrease in promoter activity between -268/-251 for
62
1.50
184hTERT
MCF-10A
Relative Expression
1.25
1.00
0.75
0.50
0.25
0.00
NRF-1
NRF-1
GABPa
GABPα
GABPb
GABPβ
BRCA1
BRCA1
Figure 3.8. Endogenous mRNA expression of GABPβ is reduced in MCF-10A cells.
Endogenous mRNA levels of NRF-1, GABPα, GABPβ, and BRCA1 were compared between
wild-type 184hTERT and MCF-10A cells using qRT-PCR. Values were normalized relative to an
HPRT internal control and expression for each gene is expressed relative to 184hTERT
expression. qRT-PCR analysis is representative of one experiment and values represent averages
of triplicate samples +/- SD.
63
184hTERT shNRF1-m1 (Figure 3.9 a) and a slightly greater than 2-fold decrease in MCF-10A
shNRF1-m1 cells (Figure 3.9 b), consistent with the loss of an NRF-1 regulatory binding site.
However, compared to the uninduced condition there was no substantial decrease in promoter
activity observed in 184hTERT shNRF1-m1 cells within 24 hours of NRF-1 knockdown, and by
72 hours the activity increased by 2.0-2.5 fold (Figure 3.9 a). This unexpected recovery of
expression despite NRF-1 depletion is consistent with the increased GABPβ RNA and protein
expression observed in Figure 3.6 and suggests regulatory compensation by another mechanism
triggered by decreased promoter activity. This maintenance of steady-state expression has already
been observed for GABPα in heterozygous knockout mice (94), suggesting a similar regulatory
mechanism is in place for GABPβ. In contrast, these constructs identified a persistent decrease in
promoter activity in MCF-10A shNRF1-m1 cells over 72 hours of NRF-1 knockdown (Figure 3.9
b). Although the decreased activity agrees with a regulatory role for NRF-1 at the GABPβ
promoter, the persistent decrease in promoter activity is not consistent with the increased protein
expression of BRCA1 and GABPβ by 72 hours (Figure 3.7). This discordance can be explained
by post-translational modifications of GABPβ. Specifically, the high molecular weight bands that
appear following 72 hr NRF-1 knockdown (Figure 3.7 a) are indicative of GABPβ
phosphorylation. As phosphorylation of GABPα and β has previously been demonstrated to
stimulate transcriptional activation of target genes including acetylcholine receptor genes [152],
this could provide an explanation for the resultant increase in BRCA1 expression (Figure 3.7 a,b).
The difference in GABPβ promoter activity response in 184hTERT and MCF-10A shNRF1-m1
cells may reflect underlying genetic differences resulting from the immortalization of the original
cell lines, and is supported by the difference in endogenous GABPβ expression already observed
with WT MCF-10A cells (Figure 3.8).
64
A.
4.5
UT
24
72
Relative Expression
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
4
-2
05
+
19
4
-2
51
+
19
4
-2
68
+
19
4
-4
78
+
19
4
-5
45
+
19
5
94
+
17
4
19
15
+
-5
-1
-7
02
3
+1
94
0.0
GABPB Promoter Construct
B.
Relative Expression
2.0
UT
24
72
1.5
1.0
0.5
4
-2
0
5+
19
4
-2
5
1+
19
4
-2
6
8+
19
4
-4
7
8+
19
4
-5
4
5+
19
5
-5
9
4+
17
4
5+
19
-7
1
-1
0
23
+1
94
0.0
GABPB Promoter Construct
Figure 3.9. GABPβ promoter activity differs between 184hTERT and MCF-10A cells upon
NRF-1 depletion. A series of 5’ GABPβ promoter deletion constructs with a Renilla expression
vector and Cmv-luciferase internal control were transfected into 184hTERT (A) and MCF10A
(B) shNRF1-m1 cells with inducible NRF-1 knockdown. Cells were induced for 24 or 72 hours
with 1ug/mL doxycycline or left uninduced (UT), and assayed for dual-luciferase activity 48
hours post-transfection. Values were normalized using the internal control, and GABPβ promoter
activity is expressed relative to the full length uninduced shNRF1-m1. Transfections were
repeated twice and values represent average results of triplicate samples +/- SD.
65
3.4.3 Loss of NRF-1 is associated with mitochondrial dysfunction in mammary epithelial cells
NRF-1 is a well characterized regulator of mitochondrial metabolism, and in conjunction
with GABPβ, coordinates the expression of several nuclear encoded genes that are essential
components of the mitochondrial respiratory chain [85,90]. To determine whether NRF-1
depletion results in mitochondrial dysfunction, we examined cell proliferation in 184hTERT and
MCF-10A shNRF1-m1 cells forced to rely on mitochondrial oxidative phosphorylation for
energy production. As in vitro cell lines typically have metabolic adaptations to generate ATP
primarily through glycolysis despite abundant oxygen supply, we circumvented this phenomenon
by substituting galactose for glucose in a reduced growth media which shifts energy production
towards mitochondrial oxidative phosphorylation [153,154]. Cells with functional mitochondria
are able to grow normally in glucose- and galactose-containing media with energy produced
through glycolysis and oxidative phosphorylation respectively. Mitochondrial dysfunction
therefore impairs the proliferative ability of cells when grown in galactose media, but not in the
presence of glucose. 184hTERT WT cells were able to grow in both glucose and galactosecontaining media, while for induced shNRF1-m1 cells (Gal-DOX) in galactose, proliferation was
reduced by 77% by day 5 in comparison to WT, with no effect in glucose-enriched media (GluDOX) (Figure 3.10 a). Uninduced shNRF1-m1 cells had an intermediate level of NRF-1
knockdown as mentioned previously, and accordingly cell proliferation decreased by 45% in
galactose-containing media (Gal-UT) compared to WT cells, with no effect on proliferation
observed in glucose-enriched media (Glu-UT). While MCF-10A WT and shNRF1-m1 cells
exhibited the same growth trend as 184hTERT WT and shNRF1-m1 cells (Figure 3.10 b), the
total number of MCF-10A cells by Day 5 was notably reduced in comparison. As MCF-10A cells
require specialized growth media containing a high concentration of glucose, the reduced growth
66
A.
2.0×10 6
Glu-WT
Glu-UT
Glu-DOX
Gal-WT
Gal-UT
Gal-DOX
Number of Cells
1.5×10 6
1.0×10 6
5.0×10 5
0
1
2
3
4
5
Day
B.
6.0×10 5
Glu-WT
Glu-UT
Glu-DOX
Gal-WT
Gal-UT
Gal-DOX
Number of Cells
4.5×10 5
3.0×10 5
1.5×10 5
0
0
1
2
3
4
5
Day
Figure 3.10. Mitochondrial dysfunction results from NRF-1 knockdown. 184hTERT (A) and
MCF-10A (B) shNRF1-m1 cells were grown in glucose- (Glu) or galactose-enriched (Gal) media,
with the number of viable cells counted over a period of 5 days. Wild type (WT), uninduced
(UT), and Dox-induced (DOX) cells represent a decreasing gradient of NRF-1 expression. Cell
proliferation experiment was performed once for each cell line, and values represent average
counts of duplicate samples +/- SD.
67
media utilized by this assay is likely limiting their rate of proliferation. Collectively, the reduced
cell growth in galactose-enriched media by uninduced and induced shNRF1-m1 cells indicates
that a mitochondrial defect results from NRF-1 knockdown.
3.4.4 The expression of genes involved in mitochondrial metabolism and apoptosis are downregulated with NRF-1 knockdown
To investigate the source of the mitochondrial dysfunction in NRF-1 deficient cells, we
examined the expression of several genes that are known NRF-1 targets and are required for
mitochondrial respiratory function. The nuclear encoded mitochondrial transcription factor A
(Tfam) is regulated by NRF-1, and is required for mitochondrial DNA (mtDNA) transcription and
replication [155]. Tfam knockout embryos have been shown to have undetectable mtDNA and no
functional respiratory chain [95]. Analysis of Tfam expression upon 72 hr induction of
184hTERT shNRF1-m1,-m2,-m6 revealed a gradient decrease of Tfam expression, with shNRF1m1 having the lowest expression (36%) and shNRF1-m6 having the highest (72%) compared to
WT cells using qRT-PCR (Figure 3.11 a). While NRF-1 expression is lowest in 184hTERT
shNRF1-m2/-m6 cells (Figure 3.2 a), the increased Tfam expression observed in shNRF1-m6
could be due to the increased expression of GABPα/β observed at 72 hours of NRF-1 knockdown
(Figure 3.6) as the Tfam promoter has also been shown to be regulated by GABP [97]. MCF-10A
shNRF1-m1,-m2,-m6 cells followed a similar trend in Tfam expression with shNRF1-m1 having
the lowest expression (34%) and shNRF1-m2/-m6 reduced by nearly 2-fold in comparison to WT
cells (Figure 3.11 b). Cytochrome c is also a direct target of NRF-1 regulation and an integral
member of the electron transport chain but also an important pro-apoptotic factor, with its release
from the mitochondria initiating the apoptotic cascade [156-158]. Induced 184hTERT and MCF10A shNRF1-m1,-m2,-m6 cells had a reduction in cytochrome c expression ranging from
68
B.
Tfam
UT
72
1.00
0.75
0.50
0.25
0.00
UT
72
1.25
1.00
0.75
0.50
0.25
6
m
R
F1
N
D.
cytochrome c
1.50
1.50
UT
72
1.25
1.00
Relative Expression
Relative Expression
2
m
MCF10A
184hTERT
C.
sh
N
sh
sh
N
R
F1
-
m
1
W
T
m
6
m
2
sh
NR
F1
-
sh
NR
F1
-
m
1
sh
NR
F1
-
W
T
0.00
R
F1
-
Relative Expression
1.25
Relative Expression
A.
0.75
0.50
0.25
UT
72
1.25
1.00
0.75
0.50
0.25
6
R
F1
-m
sh
N
2
R
F1
-m
1
sh
N
184hTERT
sh
N
W
T
R
F1
-m
6
m
2
m
0.00
sh
NR
F1
-
sh
NR
F1
-
1
m
sh
NR
F1
-
W
T
0.00
MCF10A
Figure 3.11. Decreased Tfam and cytochrome c expression is associated with NRF-1
knockdown. RNA levels of Tfam (A, B) and cytochrome c (C, D) were analyzed using qRT-PCR
in 184hTERT and MCF10A wild type (WT) and shNRF1-m1,-m2,-m6 cells over 72 hours of
1ug/mL doxycycline induction to decrease NRF-1 expression. RNA was prepared from
uninduced (UT) and 72 hour dox-induced (72) cells approximately 72 hours post-treatment.
Values were normalized relative to an HPRT internal control and fold changes are expressed
relative to a non-manipulated wild type (WT) control (relative expression set to 1.0). qRT-PCR
analysis is representative of two experiments and values represent averages of triplicate samples
+/- SD.
69
24% to 39% compared to untreated WT cells (Figure 3.11 c,d). A lack of cytochrome c in vitro
has been shown to induce respiratory chain dysfunction through the disruption of respiratory
complex assembly and stability [159]. Therefore, the decreased Tfam and cytochrome c
expression resulting from NRF-1 knockdown could be contributing to the observed defective
mitochondrial activity through respiratory chain dysfunction.
3.5 Discussion
NRF-1 is a known transcriptional regulator of mitochondrial biogenesis and function
[97], and in addition to the regulation of nuclear encoded mitochondrial proteins, it has recently
been demonstrated that NRF-1 exists in a regulatory network with GABPβ and BRCA1 [65].
Given the role of BRCA1 in mammary gland differentiation, we examined the role of NRF-1
during acinar morphogenesis through lentiviral-mediated knockdown. Our study shows that a
reduction of functional NRF-1 results in abnormal acinar formation characterized by altered cell
polariziation and an increased infilling of the central lumen (Figure 3.3-14). Acinar formation
requires both polarization of the outer cells mediated by interactions with the extra-cellular
matrix, and caspase-dependent apoptosis of the central cells [151]. Both these processes have
recently been shown to be regulated by a cyclic AMP (cAMP)-dependent Protein Kinase A
(PKA) mechanism in vitro [160]. Cell polarization was mediated by cAMP-dependent
redistribution of the α6 integrin to the cell periphery, while luminal cell hollowing accompanied a
PKA-dependent post-transcriptional induction of the proapoptotic protein Bcl-2 Interacting
Mediator (BIM), leading to cell death. Although the mechanism by which NRF-1 expression
influences cell polarization is unknown, its regulation by cAMP through activation of PGC-1α
expression [161] provides a link to the cAMP regulatory cascade. A role for NRF-1 in lumen
formation is more apparent through the probable induction of caspase-dependent apoptosis of the
70
central luminal cells. This is supported by the observed decrease in expression of the NRF-1
regulated Tfam and cytochrome c genes (Figure 3.11). In vitro studies of human cells without
mtDNA resulting from Tfam depletion demonstrate no functional oxidative phosphorylation due
to the absence of electron transport chain proteins that are encoded within the mitochondrial
genome [162]. These cells have been shown to be resistant to the pro-apoptotic effects of
staurosporine [163], demonstrating that a lack of oxidative phosphorylation resulting from
respiratory chain defects can contribute to anti-apoptotic phenotypes. Similarly, in vitro knockout
cell models of cytochrome c exhibit reduced activation of caspase-3 and a resistance to
mitochondrial induced apoptotic stimuli including UV exposure, serum deprivation and
staurosporine [164]. As respiratory chain deficiencies can contribute to impaired mitochondrial
induced apoptosis [165], the abnormal lumen formation in NRF-1 depleted acini could be due to
apoptotic resistance of the central cells resulting from respiratory chain deficiency, insufficient
releasable cytochrome c, or a combination of both.
Apoptosis is a notable feature of the developing mouse mammary terminal end bud
(TEB), with approximately 11% of cells constantly undergoing programmed cell death,
presumably to maintain a hollow luminal space [166]. Interestingly, luminal infilling is an early
hallmark of mammary epithelial cancers as it is frequently observed in the invasive carcinoma
precursor, ductal carcinoma in situ (DCIS) [151]. Luminal infilling results from decreased
apoptosis and resistance to anti-proliferative signals, which is the case for anti-apoptotic and
proliferative oncogenes [151]. We have already suggested an anti-apoptotic role for NRF-1
through down-regulation of mitochondrial oxidative phosphorylation and cytochrome c release,
and although increased cell proliferation was not observed during monolayer growth of induced
184hTERT and MCF-10A shNRF1-m1 in glucose-containing media (Figure 3.10), it is still
71
possible that these cells demonstrate some resistance to anti-proliferative signals. NRF-1 has been
shown to coordinate with E2F transcription factors to regulate a number of genes responsible for
cell proliferation, DNA replication and cell cycle progression [167]. NRF-1 knockdown in vivo
has been shown to result in a significant reduction of E2F-responsive gene expression [167],
supporting a role for NRF-1 in cell proliferation. Therefore, depletion of NRF-1 could result in
luminal infilling through a reduction of the apoptotic cascade while also initiating the aberrant
regulation of genes involved in cell proliferation.
A recent characterization of oxidative phosphorylation function in human mammary
epithelial cell lines revealed extensive mitochondrial dysfunction in tumorigenic, but not in
normal breast cells lines [168]. Similar to our results, normal MCF-10A cells were viable in both
glucose- and galactose-containing media, while tumorigenic cells including the invasive human
breast adenocarcinoma cell line MDA-MB-231 had cell growth reduced by approximately 50% in
galactose-containing media [168]. Interestingly, when human cells devoid of endogenous
mitochondria were supplemented with mitochondria from normal and tumorigenic cells, only
cells containing normal mitochondria were able to proliferate under conditions favoring oxidative
phosphorylation. As these cybrid cells contain the same nuclear DNA, the inability to grow under
galactose-containing conditions is attributed to mitochondrial defects present in the mitochondria
of tumorigenic cells. The depletion of NRF-1 in our 184hTERT and MCF-10A cell lines resulted
in growth rates similar to tumorigenic breast cell lines in galactose-containing media, supporting
the conclusion that NRF-1 is essential for normal mitochondrial function.
A unique feature of tumor cells is their predominant use of glycolysis to metabolize
glucose, which is in contrast to the mitochondrial oxidative phosphorylation performed by normal
cells; a phenomenon termed the Warburg effect [109,147]. Although this shift in metabolism has
72
been suggested to result from mitochondrial defects which perturb metabolism by oxidative
phosphorylation [109], it is also beneficial to cells by facilitating the uptake and incorporation of
nutrients into the biomass needed for rapid cell proliferation [109]. Several studies have reported
important differences between normal and tumorigenic mitochondria including aberrations in the
mitochondrial ultrastructure [169], expression of respiratory chain components [170], oxidative
phosphorylation capacity [171], and mtDNA levels [172]. Given the role of NRF-1 in
mitochondrial metabolism through its regulation of genes involved in mtDNA transcription and
replication, mitochondrial translation, and expression of respiratory chain subunits [90], aberrant
regulation and expression of NRF-1 could be contributing to the mitochondrial alterations and the
Warburg effect observed in tumor cells. Interestingly, the pharmacological drug dichloroacetate
(DCA) has been used to exploit the Warburg effect characteristic of tumor cells, as its use reverts
the metabolic shift back towards oxidative phosphorylation promoting glucose metabolism
through the mitochondria, thereby increasing apoptosis and diminishing tumor growth and
proliferation [173].
Although our results suggest a defect in cell polarization and apoptosis is the underlying
cause for the impaired acinar formation following NRF-1 depletion, the possibility of a defect in
differentiation remains plausible. Although in monolayer cell culture, NRF-1 depletion did not
have a lasting knockdown effect on BRCA1 expression (Figure 3.6, 16), further experimentation
is required to determine its influence during acinar differentiation in 3D culture. Cell
differentiation has also been shown to be modulated by mitochondrial metabolism, with the
spontaneous differentiation of human embryonic and induced-pluripotent stem cells to fibroblasts
accompanying a shift in metabolism from glycolysis to mitochondrial oxidative phosphorylation
[174]. During this differentiation process, NRF-1 and nuclear-encoded mitochondrial genes
73
including Tfam are generally up-regulated in undifferentiated cells compared to the fully
differentiated fibroblasts [174]. We have observed similar increases in NRF-1 expression during
breast cell differentiation (unpublished work). Additionally, when differentiated fibroblasts
undergo cellular reprogramming to an undifferentiated state, their metabolic state reverts back
from oxidative phosphorylation to glycolysis. Generally, it appears as though the level of
oxidative phosphorylation increases as cells become more differentiated. This may be due in part
to the production of reactive oxygen species (ROS) from oxidative phosphorylation. The
hematopoietic system has been shown to be regulated by ROS homeostasis, with stem cell
differentiation stimulated by increasing ROS levels [175]. Similarly, the differentiation of
adipocytes which accompanies increased mitochondrial metabolism is stimulated by ROS
produced by mitochondrial complex III of the respiratory chain [176]. Therefore, the extent to
which ROS can be regulated during differentiation is dependent on mitochondrial function.
Through its regulation of mitochondrial metabolism, NRF-1 may play a necessary role mediating
the regulation of metabolic state and cell differentiation.
3.6 Conclusions
Although the degree of involvement of NRF-1 in BRCA1-mediated differentiation
requires additional experimentation, our current results indicate that NRF-1 plays a role in the
formation of mammary acini in vitro as lentiviral-mediated knockdown of NRF-1 resulted in
impaired cell polarization and luminal formation. Furthermore, these cells exhibited notable
mitochondrial dysfunction when cultured in galactose. As luminal hollowing is dependent on the
maintenance of a polarized epithelial layer and selective mitochondrial-induced apoptosis of the
central luminal cells, NRF-1 may be mediating cell polarization and/or apoptosis through the
maintenance of functional mitochondria having the capacity to undergo apoptosis given its
74
regulation of genes involved in mitochondrial metabolism and function including Tfam and
cytochrome c.
3.7 Acknowledgements
We would like to thank Rachael Klinoski for her excellent technical assistance. This work was
funded by a grant from the Canadian Breast Cancer Foundation – Ontario Region.
75
CHAPTER 4 : General Discussion
4.1
Metabolic Adaptations in Tumors
Among the many characteristics acquired by tumor cells is an altered metabolism
characterized by a shift from mitochondrial oxidative phosphorylation towards aerobic glycolysis
for the generation of ATP; a phenomenon termed the Warburg effect [107]. This increased
dependency on glycolysis for ATP is present in most human cancers, and forms the basis for
tumor imaging using 18fluorodeoxyglucose positron-emission tomography [177]. It also provides
a growth advantage for tumor cells for a number of reasons (summarized in [178]). First, a
reliance on glycolysis allows tumor cells to survive under the hypoxic conditions resulting from
inadequate tumor vasculature [179], where a dependence on oxidative phosphorylation would
result in tumor lethality. Second, increased acid in the tumor microenvironment resulting from
lactic acid production through glycolysis has been suggested to promote tumor invasion [180],
and suppress the cytotoxic activity of T cells [181]. Furthermore, the released lactic acid from
tumor cells can be recycled into pyruvate by surrounding stromal cells through complementary
metabolic processes that exist between stromal cells and the surrounding carcinomas [182]. This
relationship allows for continual regeneration of pyruvate to be used by tumor cells during
oxidative phosphorylation or lactic acid fermentation, allowing for prolonged tumor growth and
survival. Third, intermediates from glycolysis are utilized as substrates in anabolic reactions
conferring cell growth and proliferation including the pentose phosphate pathway, and the
synthesis of lipids, amino acids, and nucleic acids [183].
4.2
Metabolism and Differentiation of Embryonic Stem Cells
Pluripotent embryonic stem cells (ESC) have the capacity to both self-renew and
76
differentiate to form the germ layers during embryogenesis. Recent studies have determined that
mouse ESCs are able to switch between oxidative phosphorylation and glycolysis, while the more
differentiated form of ESCs, epiblast stem cells (EpiSC) which are similar to human ESC, are
reliant primarily on glycolysis [184]. Although EpiSC/hESCs had more developed mitochondria
and higher mtDNA copy numbers than ESCs, consistent with increased PGC-1α expression, the
mitochondrial function was decreased in comparison to ESCs. The source of the muted
mitochondrial function was attributed to decreased complex IV cytochrome c oxidase (COX)
activity, with 20 out of its 22 nuclear-encoded genes downregulated. Decreased expression of
regulators of mitochondrial metabolism including the co-regulator PGC-1β and an ERR
transcription factor was observed in EpiSC/hESC but not ESC, suggesting a source for the
mitochondrial dysfunction. Notably, a series of GABP binding sites exist with the COX IV
promoter [90], and NRF-1 has been shown to coordinate the transcriptional regulation of all ten
nuclear encoded COX subunits, including COX IV, in neurons [185]. Hypoxia Inducible Factor-1
alpha (HIF-1α) was identified as a regulator of metabolic reprogramming during ESC
differentiation, where its overexpression induced a shift in metabolism from oxidative
phosphorylation towards glycolysis. This finding is consistent with its role in promoting
glycolysis under hypoxic conditions [186] while inhibiting mitochondrial biogenesis through the
negative regulation of c-Myc dependent PGC-1β expression [187]. Metabolic regulation of stem
cell diferentiation might also be present in the mouse mammary gland as HIF-1α has been
previously identified as a critical regulator of differentiation and lipid secretion [188].
Interestingly, HIF-1α and c-Myc have already been shown to play a similar role regulating the
glycolytic shift in tumors, suggesting that cancer cells exploit the metabolic reprogramming
process of ESCs in order to facilitate their rapid growth.
77
4.3
Metabolic Reprogramming in Tumourigenesis
An emerging view of tumourigenesis suggests that alterations in cancer metabolism are
contributing factors to the classical hallmarks of cancer including continual growth, evasion of
growth suppression, resisting apoptosis, limitless replicative potential, sustained angiogenesis,
and initiation of invasion and metastasis [178,189]. In addition to the activation of HIF-1,
modulation of p53 and NF-κB activity which regulate genes involved in suppressing glycolysis
while increasing oxidative phosphorylation might also contribute to metabolic reprogramming
[183] and lead to self-sufficient growth, limitless replication and evasion of growth suppression
[189]. In mice, NF-κB is able to regulate oxidative phosphorylation through activation of NRF-1
when coupled with CREB [190], while also able to upregulate HIF-1α expression in response to
hypoxia [191]. Mitochondrial dysfunction has also been shown to be sufficient in inhibiting p53
expression and function [192], and dysregulation of oxidative phosphorylation can contribute to
apoptotic resistance while byproducts of the metabolic process influence tumor invasion,
metastasis and immunosuppression.
It is also possible that the metabolic shift and hallmarks of cancer arise independently of
one another, with their co-evolution resulting from the selective pressure for rapidly proliferating
cells to evade multiple tumor suppressive processes [178]. Regardless, the common involvement
of mitochondrial metabolism and a metabolic shift in tumors has several therapeutic implications.
Increased understanding of the mechanisms driving the metabolic reprogramming in tumor cells
could be used to inhibit the glycolytic switch, thereby impairing cell transformation and tumor
development. As metabolism has also been shown to regulate stem cell differentiation, insight
into the networks controlling metabolic-mediated differentiation could lead to new insights into
78
cancer stem cell differentiation and generate novel therapies to target this elusive subpopulation
of tumor-initiating cells.
79
CHAPTER 5 : Summary and Conclusions
NRF-1 forms a complex with several unidentified proteins to regulate the GABPβ
promoter. We have attempted to identify these co-activators through the use of an immobilizedtemplate assay with the GABPβ regulatory site. However, the presence of non-specific DNA
binding proteins in the isolated fractions prevented the identification of co-activators. Through its
interaction with GABPβ, NRF-1 is potentially involved in BRCA1-mediated cellular
differentiation. Using an in vitro 3D model of mammary gland development, knockdown of NRF1 results in impaired mammary differentiation characterized by a lack of cell polarization and
luminal infilling. Although monolayer culture of NRF-1 deficient cells does not result in a
sustained downregulation of GABPβ/BRCA1 expression, it may play a role during differentiation
when the NRF-1 pathway seems to be more significant. Additionally, these cells display some
mitochondrial dysfunction and downregulation of known NRF-1 target genes involved in
mitochondrial transcription, the respiratory chain and apoptosis. As NRF-1 is a regulator of
nuclear-encoded mitochondrial proteins involved in mitochondrial biogenesis and metabolism, its
downregulation could possibly result in a defect in oxidative phosphorylation or mitochondrialdependent apoptosis that would be reflected in impaired mammary gland formation. Further work
is required to investigate the role for NRF-1 in differentiation, as a metabolic switch between
oxidative phosphorylation and glycolysis has been shown to regulate stem cell differentiation. As
this same type of switch is frequently observed in cancer, NRF-1 and mitochondrial activity may
also be involved in altered tumor cell metabolism.
80
Reference List
1. Canadian Cancer Society's Steering Committee on Cancer Statistics. Canadian Cancer
Statistics. Toronto, ON : Canadian Cancer Society, 2012.
2. Millikan RC, Newman B, Tse CK, Moorman PG, Conway K, Dressler LG, et al.. 2008,
Epidemiology of basal-like breast cancer, Breast Cancer Res Treat. 109:123-139.
3. Carey LA, Perou CM, Livasy CA, Dressler LG, Cowan D, Conway K, et al. 2006, Race,
breast cancer subtypes, and survival in the Carolina Breast Cancer Study, JAMA. 295:2492-502.
4. Perou CM, Sørlie T, Eisen MB, van de Rijn M, Jeffrey SS, Rees CA, et al. 2000, Molecular
portraits of human breast tumours. Nature. 406:747-752.
5. Rouzier R, Perou CM, Symmans WF, Ibrahim N, Cristofanilli M, Anderson K, et al.
2005, Breast cancer molecular subtypes respond differently to preoperative chemotherapy.
Clinical Cancer Research. 11:5678-5685.
6. Visvader, JE. . 2009, Keeping abreast of the mammary epithelial hierarchy and breast
tumourigenesis, Genes and Development. 23:2563-2577.
7. Yu K, Lee CH, Tan PH, Tan P. 2004, Conservation of breast cancer molecular subtypes and
transcriptional patterns of tumor progression across distinct ethnic populations. Clin Cancer Res,
10:5508-5517.
8. Muggerud AA, Hallett M, Johnsen H, Kleivi K, Zhou W, Tahmasebpoor S, Amini RM,
Botling J, Børresen-Dale AL, Sørlie T, Wärnberg F. 2010, Molecular diversity in ductal
carcinoma in situ (DCIS) and early invasive breast cancer. Molec Oncol. 4:357-368.
9. Curtis C, Shah SP, Chin SF, Turashvili G, Rueda OM, Dunning MJ, et al. 2012, The
genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups. Nature,
486:346-352
10. Lineman GJ, Visvader JE. 2010, Insights into the cell of origin in breast cancer and breast
cancer stem cells. Asia Pac J Clin Oncol. 6;89-97.
11. Sims AH, Howell A, Howell SJ, Clarke RB. 2007, Origins of breast cancer subtypes and
therapeutic implications. Nature Clin Prac Oncol. 4:516-525.
12. Johnson, MC. Anatomy and Physiology of the Breast. [book auth.] M. Kaufmann I. Jatoi.
Management of Breast Diseases. Berlin Heidelbery : Springer, 2010.
13. Radisky E, Radisky D. 2007, Stromal induction of breast cancer: Inflammation and
invasion. Rev Endo Met Disorders. 8:279-287.
81
14. Hinck L, Silberstein G. 2005, Key stages in mammary gland development: The mammary
end bud as a motile organ. Breast Cancer Res. 7:245-251.
15. Wilde C, Knight C, Flint D. 1999, Control of milk secretion and apoptosis during mammary
involution. J of Mammary Gland Biol Neoplasia. 4:129-136.
16. Baxter F, Neoh K, Tevendale M. 2007, The beginning of the end: death signals in early
involution. J Mammary Gland Biol Neoplasia. 12:3-13.
17. Stingl J, Eaves C, Zandieh I, Emerman J. 2001, Characterization of bipotent mammary
epithelial progenitor cells in normal adult human breast tissue. Breast Cancer Res Treat. 67:93109.
18. Dontu G, Abdallah W, Foley J, Jackson K, Clarke M, Kawamura M, et al. 2003, In vitro
propagation and transcriptional profiling of human mammary stem/progenitor cells. Genes Dev,
15:1253-1270.
19. Villadsen R, Fridriksdottir A, Rønnov-Jessen L, Gudjonsson T, Rank F, LaBarge M, et
al. 2007, Evidence for a stem cell hierarchy in the adult human breast. J Cell Biol. 177:87-101.
20. Tsai Y, Lu Y, Nichols P, Zlotnikov G, Jones P, Smith H. 1996, Contiguous patches of
normal human mammary epithelium derived from a single stem cell: implications for breast
carcinogenesis. Cancer Res. 56:402-404.
21. Shackleton M, Vaillant F, Simpson K, Smyth G, Asselin-Labat M, Wu L, et al. 2006,
Generation of a functional mammary gland from a single stem cell. Nature. 5:84-88.
22. Lehnert B, Iyer R. 2002, Exposure to low-level chemicals and ionizing radiation: reactive
oxygen species and cellular pathways. Hum Exp Toxicol. 21:65-69.
23. Ohshima H, Tatemichi M, Sawa T. 2003, Chemical basis of inflammation-induced
carcinogenesis. Arch Biochem Biophys. 417:3–11.
24. Hausen H. 2000, Papillomaviruses Causing Cancer: Evasion From Host-Cell Control in
Early Events in Carcinogenesis. J Nat Cancer Inst. 92:690-698.
25. Rossi D, Seita J, Czechowicz A, Bhattacharya D, Bryder D, Weissman I. 2007,
Hematopoietic stem cell quiescence attenuates DNA damage response and permits DNA damage
accumulation during aging. Cell Cycle. 6:2371-2376.
26. Reya T, Morrison S, Clarke M, Weissman I. 2001, Stem cells, cancer, and cancer stem
cells. Nature. 414:105-111.
82
27. Ponti D, Zaffaroni N, Capelli C, Daidone MG. 2006, Breast cancer stem cells: An
overview. Eur J Cancer. 42:1219-1224.
28. Tan BT, Park CY, Ailles LE, Weissman IL. 2006, The cancer stem cell hypothesis: a work
in progress. Lab Inves. 86:1203-1207.
29. Ricardo S, Vieira AF, Gerhard R, Leitão D, Pinto R, Cameselle-Teijeiro JF, et al. 2011,
Breast cancer stem cell markers CD44, CD24 and ALDH1: expression distribution within
intrinsic molecular subtype. J Clin Path. 64:937-946.
30. Al-Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ, Clarke MF. 2003, Prospective
identification of tumorigenic breast cancer cells. PNAS. 100:3983-3988.
31. Honeth G, Bendahl P, Ringnér M, Saal LH, Gruvberger-Saal SK, Lövgren K, et al.
2008, The CD44+/CD24- phenotype is enriched in basal-like breast tumors. Breast Cancer Res.
10:R53-R65.
32. Sorlie T, Tibshirani R, Parker J, Hastie T, Marron JS, Nobel A, et al. 2003, Repeated
observation of breast tumor subtypes in independent gene expression data sets. PNAS USA. 100:
8418-8423.
33. Melchor L, Benítez J. 2008, An integrative hypothesis about the origin and development of
sporadic and familial breast cancer subtypes. Carcinogenesis. 29:1475-1482.
34. Dontu G, Al-Hajj M, Abdallah WM, Clarke MF, Wicha MS. 2003, Stem cells in normal
breast development and breast cancer. Cell Prolifl. 36:Suppl 1, 59-72.
35. Molyneux G, Geyer FC, Magnay FA, McCarthy A, Kendrick H, Natrajan R, Mackay A,
Grigoriadis A, Tutt A, Ashworth A, Reis-Filho JS, Smalley MJ. 2010, BRCA1 basal-like
breast cancers originate from luminal epithelial progenitors and not from basal stem cells. Cell
Stem Cell. 3:403-417.
36. Lim E, Vaillant F, Wu D, Forrest NC, Pal B, Hart AH, Asselin-Labat ML, Gyorki DE,
Ward T, Partanen A, et al. 2009, Aberrant luminal progenitors as the candidate target
population for basal tumor development in BRCA1 mutation carriers. Nature Med. 15:907-913.
37. Proia TA, Keller PJ, Gupta PB, Klebba I, Jones AD, Sedic M, et al. 2011, Genetic
Predisposition Directs Breast Cancer Phenotype by Dictating Progenitor Cell Fate. Cancer Stem
Cell. 8:149–163.
83
38. Kima J, Villadsena R, Sørlieb T, Fogha L, Grønlunda SZ, Fridriksdottira AJ, et al.
2012, Tumor initiating but differentiated luminal-like breast cancer cells are highly invasive in
the absence of basal-like activity. PNAS. 10.1073/pnas.1203203109
39. Kenemans P, Verstraeten RA, Verheijen RHM. 2004, Oncogenic pathways in hereditary
and sporadic breast cancer. Maturitas. 49:34–43.
40. Berger AH, Knudson AG, Paolo Pandolfi PP. 2011, A continuum model for tumour
suppression. Nature. 476:163–169.
41. Konishi H, Mohseni M, Tamaki A, Garay JP, Croessmann S, Karnan S, et al. 2011,
Mutation of a single allele of the cancer susceptibility gene BRCA1 leads to genomic instability
in human breast epithelial cells. PNAS. 108:17773-17778.
42. Marchina E, Fontana MG, Speziani M, Salvi A, Ricca G, Di Lorenzo D, et al. 2010,
BRCA1 and BRCA2 genetic test in high risk patients and families: counselling and management.
Oncol Rep. 24:1661-1667.
43. Taylor J, Lymboura M, Pace PE, A'Hern RP, Desai AJ, Shousha S, et al. 1998, An
important role for BRCA1 in breast cancer progression is indicated by its loss in a large
proportion of non-familial breast cancers. Int J Cancer. 79:334-342.
44. Honrado E, Benítez J, Palacios J. 2005, The molecular pathology of hereditary breast
cancer: genetic testing and therapeutic implications. Mod Pathol. 18:1305-1320.
45. Atlas E, Stramwasser M, Whiskin K, Mueller CR. 2000, GA-binding protein alpha/beta is
a critical regulator of the BRCA1 promoter. Oncogene. 19:1933-1940.
46. Mancini DN, Rodenhiser DI, Ainsworth PJ, O'Malley FP, Singh SM, Xing W, et al. 9,
1998, CpG methylation within the 5' regulatory region of the BRCA1 gene is tumor specific and
includes a putative CREB binding site. Oncogene. 16:1161-1169.
47. Catteau A, Harris WH, Xu CF, Solomon E. 1999, Methylation of the BRCA1 promoter
region in sporadic breast and ovarian cancer: correlation with disease characteristics. Oncogene.
18:1957-1965.
48. Xu X, Weaver Z, Linke SP, Li C, Gotay J, Wang XW, Harris CC, Ried T, Deng CX.
1999, Centrosome amplification and a defective G2-M cell cycle checkpoint induce genetic
instability in BRCA1 exon 11 isoform-deficient cells. Mol Cell. 3:389-395.
84
49. Vaughn JP, Davis PL, Jarboe MD, Huper G, Evans AC, Wiseman RW, et al. 1996,
BRCA1 expression is induced before DNA synthesis in both normal and tumor-derived breast
cells. Cell Growth and Differ. 7:711-715.
50. Yarden RI, Pardo-Reoyo S, Sgagias M, Cowan KH, Brody LC. 2002, BRCA1 regulates
the G2/M checkpoint by activating Chk1 kinase upon DNA damage. Nature Genet. 30:285-289.
51. Deng CX, Brodie SG. 2000, Roles of BRCA1 and its interacting proteins. BioEssays.
22:728-737.
52. Deng CX. 2001, Tumourigenesis as a consequence of genetic instability in Brca1 mutant
mice. Mutat Res. 477:183-189.
53. Weaver Z, Montagna C, Xu X, Howard T, Gadina M, Brodie SG, et al. 2002, Mammary
tumors in mice conditionally mutant for Brca1 exhibit gross genomic instability and centrosome
amplification yet display a recurring distribution of genomic imbalances that is similar to human
breast cancer. Oncogene. 21:5097-5107.
54. Brodie SG, Xu X, Qiao W, Li WM, Cao L, Deng CX. 2001, Multiple genetic changes are
associated with mammary tumorigenesis in Brca1 conditional knockout mice. Oncogene.
20:7514-7523.
55. Mueller CR, Roskelley CD. 2003, Regulation of BRCA1 and its relationship to sporadic
breast cancer. Breast Cancer Res. 5:45-53.
56. Marquis ST, Rajan JV, Wynshaw-Boris A, Xu J, Yin GY, Abel KJ, et al. 1995, The
developmental pattern of Brca1 expression implies a role in differentiation of the breast and other
tissues. Nature Genet. 11:17 - 26.
57. Medina D. 2004, Breast cancer: the protective effect of pregnancy. Clin Cancer Res.
10:380S-384S.
58. Shen SX, Weaver Z, Xu X, Li C, Weinstein M, Chen L, Guan XY, Ried T, Deng CX.
1998, A targeted disruption of the murine Brca1 gene causes gamma-irradiation hypersensitivity
and genetic instability. Oncogene. 17:3115-3124.
59. Shakya R, Reid LJ, Reczek CR, Cole F, Egli D, Lin CS, et al. 2011, BRCA1 Tumor
Suppression Depends on BRCT Phosphoprotein Binding, But Not Its E3 Ligase Activity.
Science. 334:525-528.
85
60. Hohenstein P, Kielman MF, Breukel C, Bennett M, Wiseman R, Krimpenfort P, et al.
2001, A targeted mouse Brca1 mutation removing the last BRCT repeat results in apoptosis and
embryonic lethality at the headfold stage. Oncogene. 20:2544-2550.
61. Xu X, Wagner KU, Larson D, Weaver Z, Li C, Ried T, et al. 1999, Conditional mutation
of Brca1 in mammary epithelial cells results in blunted ductal morphogenesis and tumour
formation. Nature Genet. 22:37-43.
62. Kubista M, Rosner M, Kubista E, Bernaschek G, Hengtschlager M. 2001, Brca1
regulates in vitro differentiation of mammary epithelial cells. Oncogene. 21:4747-4756.
63. Furuta S, Jiang X, Gu B, Cheng E, Chen PL, Lee WH. 2005, Depletion of BRCA1 impairs
differentiation but enhances proliferation of mammary epithelial cells. PNAS. 102:9176–9181.
64. Liu S, Ginestier C, Charafe-Jauffret E, Foco H, Kleer CG, Merajver SD, Dontu G,
Wicha MS. 2008, BRCA1 regulates human mammary stem/progenitor cell fate. PNAS.
105:1680-1685.
65. Thompson C, MacDonald G, Mueller CR. 2011, Decreased expression of BRCA1 in SKBR-3 cells is the result of aberrant activation of the GABP Beta promoter by an NRF-1containing complex. Mol Cancer. 10:62-79.
66. Ritter HD, Antonova L, Mueller CR. 2012, The unliganded glucocorticoid receptor
positively regulates the tumor suppressor gene BRCA1 through GABP beta. Mol Cancer Res. 10:
558.
67. Antonova L, Aronson K, Mueller CR. 2011, Stress and breast cancer: from epidemiology to
molecular biology. Breast Cancer Res. 13:208 .
68. Atlas E, Stramwasser M, Mueller CR. 2001, A CREB site in the BRCA1 proximal
promoter acts as a constitutive transcriptional element. Oncogene. 20:7110-7114.
69. Graves ML, Zhou L, MacDonald G, Mueller CR, Roskelley CD. 2007, Regulation of the
BRCA1 promoter in ovarian surface epithelial cells and ovarian carcinoma cells. FEBS Lett.
581:1825-1833.
70. Rauch T, Zhong X, Pfeifer GP, Xu X. 2005, 53BP1 is a positive regulator of the BRCA1
promoter. Cell Cycle. 4:1078-1083.
71. Corkery D, Thillainadesan G, Coughlan N, Mohan RD, Isovic M, Tini M, Torchia J.
2011, Regulation of the BRCA1 gene by an SRC3/53BP1 complex. BMC Biochem. 12:50.
86
72. MacLachlan TK, Dash BC, Dicker DT, El-Deiry WS. 2000, Repression of BRCA1 through
a feedback loop involving p53. J Biol Chem. 275:31869-31875.
73. Oberley MJ, Inman DR, Farnham PJ. 2003, E2F6 negatively regulates BRCA1 in human
cancer cells without methylation of histone H3 on lysine. J Biol Chem. 278:42466-42476.
74. Bindra RS, Glazer PM. 2006, Basal repression of BRCA1 by multiple E2Fs and pocket
proteins at adjacent E2F sites. Cancer Biol Ther. 5:1400-1407.
75. Wang A, Schneider-Broussard R, Kumar AP, MacLeod MC, Johnson DG. 2000,
Regulation of BRCA1 expression by the Rb-E2F pathway. J Biol Chem. 275:4532-4536.
76. Jeffy BD, Hockings JK, Kemp MQ, Morgan SS, Hager JA, Beliakoff J, et al. 2005, An
estrogen receptor-alpha/p300 complex activates the BRCA-1 promoter at an AP-1 site that binds
Jun/Fos transcription factors: repressive effects of p53 on BRCA-1 transcription. Neoplasia.
7:873-882.
77. Gorski JJ, Kennedy RD, Hosey AM, Harkin DP. 2009, The complex relationship between
BRCA1 and ERalpha in hereditary breast cancer. Clin Cancer Res. 15:1514-1518.
78. King MC, Marks JH, Mandell JB, The New York Breast Cancer Study Group. 2003,
Breast and ovarian cancer risks due to inherited mutations in BRCA1 and BRCA2. Science.
302:643-646 .
79. Thompson ME, Jensen RA, Obermiller PS, Page DL, Holt JT. 1995, Decreased
expression of BRCA1 accelerates growth and is often present during sporadic breast cancer
progression. Nature Genet. 9:444 - 450.
80. Russell PA, Pharoah PDP, De Foy K, Ramus SJ, Symmonds I, Scott I, et al. 2000,
Frequent loss of BRCA1 mRNA and protein expression in sporadic ovarian cancers.
Carcinogenesis. 87:317-321.
81. Chen Y, Xu J, Borowicz S, Collins C, Huo D, Olopade OI. 2011, c-Myc activates BRCA1
gene expression through distal promoter elements in breast cancer cells. BMC Cancer. 11:246.
82. Moskwa P, Buffa FM, Pan Y, Panchakshari R, Gottipati P, Muschel RJ, et al. 2011,
miR-182-mediated downregulation of BRCA1 impacts DNA repair and sensitivity to PARP
inhibitors. Mol Cell. 41:210-220.
83. Heyn H, Engelmann M, Schreek S, Ahrens P, Lehmann U, Krelpe H, et al. 2011,
MicroRNA miR-335 is crucial for the BRCA1 regulatory cascade in breast cancer development.
Int J Cancer. 129:2797-2806.
87
84. Chang S, Wang RH, Akagi K, Kim KA, Martin BK, Cavallone L, et al. 2011, Tumor
suppressor BRCA1 epigenetically controls oncogenic microRNA-155. Nature Med. 17:12751283.
85. Rosmarin AG, Resendes KK, Yang Z, McMillan JN, Fleming SL. 2004, GA-binding
protein transcription factor: a review of GABP as an integrator of intracellular signaling and
protein-protein interactions. Blood Cells Mols Dis. 32:143-154.
86. Batchelor AH, Piper DE, de la Brousse FC, McKnight SL, Wolberger C. 1998, The
structure of GABPalpha/beta: An ETS domain-ankyrin repeat heterodimer bound to DNA.
Science. 279:1037-1040.
87. Mackereth CD, Schärpf M, Gentile LN, MacIntosh SE, Slupsky CM, McIntosh LP.
2004, Diversity in structure and function of the Ets family PNT domains. J Mol Biol. 342:12491264.
88. Gugneja S, Virbasius JV, Scarpulla RC. 1995, Four structurally distinct non-DNA binding
subunits of human nuclear respiratory factor 2 share a conserved transcriptional activation
domain. Mol Cell Biol. 15:102-111.
89. Sawa C, Goto M, Suzuki F, Watanabe H, Sawada J, Handa H. 1996, Functional domains
of transcription factor hGABP beta/E4TF1-53 required for nuclear localization and transcription
activation. Nucleic Acids Res. 24:4954-4961.
90. Scarpulla, RC. 2008, Transcriptional paradigms in mammalian mitochondrial biogenesis and
function. Physiol Rev. 88:611-638.
91. Yang ZF, Mott S, Rosmarin AG. 2007, The Ets transcription factor GABP is required for
cell-cycle progression. Nature Cell Biol. 9:339-346.
92. Yu S, Cui K, Jothi R, Zhao DM, Jing X, Zhao K, Xu HH. 2011, GABP controls a critical
transcription regulatory module that is essential for maintenance and differentiation of
hematopoietic stem/progenitor cells. Blood. 117:2166-2178.
93. Kinoshita K, Ura H, Akagi T, Usuda M, Koide H, Yokota T. 2007, GABP alpha regulates
Oct-3/4 expression in mouse embryonic stem cells. Biochem Biophys Res Comm. 353:686-691.
94. Ristevski S, O'Leary DA, Thornell AP, Owen MJ, Kola I, Hertzog PJ. 2004, The ETS
transcription factor GABPalpha is essential for early embryogenesis. Mol Cell Biol. 24:58445849.
88
95. Larsson NG, Wang J, Wilhelmsson H, Oldfors A, Rustin P, Lewandoski M, Barsh GS,
Clayton DA. 1998, Mitochondrial transcription factor A is necessary for mtDNA maintenance
and embryogenesis in mice. Nature Genet. 18:231-236.
96. Strachan T, Read AP. Organization of the human genome. Human Molecular Genetics. 2.
s.l. : Wiley-Liss, 1999, 7.
97. Scarpulla, RC. 2002, Nuclear activators and coactivators in mammalian mitochrondrial
biogenesis. Biochim Biophys Acta. 1576:1-14.
98. Scarpulla, RC. 2011, Metabolic control of mitochondrial biogenesis through the PGC-1
family regulatory network. Biochim Biophys Acta. 1813:1269-1278.
99. Puigserver P, Spiegelman BM. 2003, Peroxisome proliferator-activated receptor-gamma
coactivator 1 alpha (PGC-1 alpha): transcriptional coactivator and metabolic regulator. Endocr
Rev. 24:78-90.
100. Ongwijitwat S, Liang HL, Graboyes EM, Wong-Riley MTT. 2006, Nuclear respiratory
factor 2 senses changing cellular energy demands and its silencing down-regulates cytochrome
oxidase and other target gene mRNAs. Gene. 374:39-49.
101. Martin ME, Chinenov Y, Yu M, Schmidt TK, Yang XY. 1996, Redox Regulation of GAbinding Protein-α DNA Binding Activity. J Biol Chem. 271:25617-25623.
102. Gugneja S, Virbasius CM, Scarpulla RC. 1996, Nuclear respiratory factors 1 and 2 utilize
similar glutamine-containing clusters of hydrophobic residues to activate transcription. Mol Cell
Biol. 16:5708–5716.
103. Gómez-Cuadrado A, Martín M, Noël M, Ruiz-Carrillo A. 1995, Initiation binding
repressor, a factor that binds to the transcription initiation site of the histone h5 gene, is a
glycosylated member of a family of cell growth regulators. Mol Cell Biol. 15:6670-6685
104. Wu Z, Puigserver P, Andersson U, Zhang C, Adelmant G, Mootha V, Troy A, Cinti S,
Lowell B, Scarpulla RC, Spiegelman BM. 1999, Mechanisms controlling mitochondrial
biogenesis and respiration through the thermogenic coactivator PGC-1. Cell. 98:115-124.
105. Huo L, Scarpulla RC. 2001, Mitochondrial DNA instability and peri-implantation lethality
associated with targeted disruption of Nuclear Respiratory Factor 1 in mice. Mol Cell Biol.
21:644–654.
106. Kim JW, Dang CV. 2006, Cancer's molecular sweet tooth and the Warburg effect. Cancer
Res. 66: 8927-8930.
89
107. Warburg O. 1956, On respiratory impairment in cancer cells. Science. 124:269-270.
108. Chen Z, Lu W, Garcia-Prieto C, Huang P. 2007, The Warburg effect and its cancer
therapeutic implications. J Bioenerg Biomembr. 39:267-274.
109. Vander Heiden MG, Cantley LC, Thompson CB. 2009, Understanding the Warburg
effect: the metabolic requirements of cell proliferation. Science. 324:1029-1033.
110. Carew JS, Huang P. 2002, Mitochondrial defects in cancer. Mol Cancer. 1:9.
111. Pelicano H, Xu RH, Du M, Feng L, Sasaki R, Carew JS, Hu Y, Ramdas L, Hu L,
Keating MJ, Zhang W, Plunkett W, 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.
112. Milne RL, Antoniou AC. 2011, Genetic modifiers of cancer risk for BRCA1 and BRCA2
mutation carriers. Ann Oncol. 22:(suppl 1)i11-i17.
113. van der Kolk DM, de Bock GH, Leegte BK, Schaapveld M, Mourits MJ, de Vires J, et
al. 2010, Penetrance of breast cancer, ovarian cancer and contralateral breast cancer in BRCA1
and BRCA2 families: high cancer indcidence at older age. Cancer Res Treat. 12:643-651.
114. Yang X, Lippman ME. 1999, BRCA1 and BRCA2 in breast cancer. Breast Cancer Res
Treat. 54:1-10.
115. Arver B, Du Q, Chen J, Luo L, Lindblom A. 2000, Hereditary breast cancer: a review.
Sem Cancer Biol. 10:271-288.
116. De Leeneer K, Coene I, Crombez B, Simkens J, Van den Broecke R, Bols A, Stragier B,
Vanhoutte I, De Paepe A, Poppe B, Claes K. 2012, Prevalence of BRCA1/2 mutations in
sporadic breast/ovarian cancer patients and identification of a novel de novo BRCA1 mutation in
a patient diagnosed with late onset breast and ovarian cancer: implications for genetic testing.
Breast Cancer Res Treat. 13:87-95.
117. Hu Z, Fan C, Oh DS, MArron JS, He X, Qaqish BF et al. 2007, The molecular portraits
of breast tumors are conserved across microarray platforms. BMC Genomics. 6:96.
118. Turner NC, Reis-Filho JS, Russell AM, Springall RJ, Ryder K, Steele D, et al. 2007,
BRCA1 dysfunction in sporadic basal-like breast cancer . Oncogene. 26:2126-2132.
119. Xu CF, Chambers JA, Solomon E. 1997, Complex regulation of the BRCA1 gene. J Biol
Chem. 272:20994-20997.
90
120. Chinenov Y, Henzl M, Martin ME. 2000, The α and β subunits of the GA-binding protein
form a stable heterodimer in solution. J Biol Chem. 275:7749-7756.
121. Xu YZ, Heravi M, Thuraisingam T, Di Marco S, Muanza T, Radzioch D. 2010, Brg-1
mediates the constitutive and fenretinide-induced expression of SPARC in mammary carcinoma
cells via its interaction with transcription factor Sp1. Mol Cancer. 9:210-227.
122. Moll JR, Acharya A, Gal J, Mir AA, Vinsona C. 2002, Magnesium is required for specific
DNA binding of the CREB B-ZIP domain. Nucleic Acids Res. 30:1240–1246.
123. . Carey MF, Peterson CL, Smale ST. 2012, Experimental Strategies for the Identification
of DNA-Binding Proteinss.l. : Cold Spring Harbor Laboratory Press, Cold Spring Harbor
Protocols, pp. 18-33.
124. Jensen, JH. 2008, Calculating pH and salt dependence of protein-protein binding. Curr
Pharm Biotech. 9:96-102.
125. Torigoe T, Izumi H, Yoshida Y, Ishiguchi H, Okamoto T, Itoh H, Kohno K. 2003, Low
pH enhances Sp1 DNA binding activity and interaction with TBP. Nucleic Acids Res. 31:45234530.
126. Evans MJ, Scarpulla RC. 1990, NRF-1: a trans-activator of nuclear-encoded respiratory
genes in animal cells. Genes Dev. 4:1023-1034.
127. Medina PP, Sanchez-Cespedes M. 2008, Involvement of the chromatin-remodeling factor
BRG1/SMARCA4 in human cancer. Epigenetics. 3: 64-68.
128. Kwon CS, Wagner D. 2007, Unwinding chromatin for development and growth: a few
genes at a time. TIG. 23:403-412.
129. Reisman D, Glaros S, Thompson EA. 2009, The SWI/SNF complex and cancer.
Oncogene. 28:1653-1668.
130. Patsialou A, Wilsker D, Moran E. 2005, DNA-binding properties of ARID family
proteins. Nucleic Acids Res. 33:66-80.
131. Dallas PB, Pacchione S, Wilsker D, Bowrin V, Kobayashi R, Moran E. 2000, The human
SWI-SNF complex protein p270 is an ARID family member with non-sequence-specific DNA
binding activity. Mol Cell Biol. 20:3137-3146.
132. Guan B, Wang TL, Shih IeM. 2011, ARID1A, a factor that promotes formation of
SWI/SNF-mediated chromatin remodeling, is a tumor suppressor in gynecologic cancers. Cancer
Res. 71:6718-6727.
91
133. Trotter KW, Fan H, Ivey ML, Kingston RE, Archer TK. 2008, The HSA domain of
BRG1 mediates critical interactions required for glucocorticoid receptor-dependent
transcriptional activation in vivo. Mol Cell Biol. 28:1413–1426.
134. Shao D, Liu Y, Liu X, Zhu L, Cui Y, Cui A, et al. 2010, PGC-1 beta-regulated
mitochondrial biogenesis and function in myotubes is mediated by NRF-1 and ERR alpha.
Mitochondrion. 10:516–527.
135. Vercauteren K, Pasko RA, Gleyzer N, Marino VM, Scarpulla RC. 2006, PGC-1-Related
Coactivator: Immediate early expression and characterization of a CREB/NRF-1 binding domain
associated with cytochrome c promoter occupancy and respiratory growth. Mol Cell Biol.
26:7409–7419.
136. Watkins G, Douglas-Jones A, Mansel RE, Jiang WG. 2004, The localization and
reduction of nuclear staining of PPARg and PGC-1 in human breast cancer. Oncol Rep. 12:483–
488.
137. Jiang WG, Douglas-Jones A, Mansel RE. 2003, Expression of peroxisome-proliferator
activated receptor-gamma (PPARgamma) and the PPARgamma co-activator, PGC-1, in human
breast cancer correlates with clinical outcomes. Int J Cancer. 106:752-757.
138. Feilchenfeldt J, Bru¨ndler MA, Soravia C, To¨tsch M and Meier CA. 2004, Peroxisome
proliferator-activated receptors (PPARs) and associated transcription factors in colon cancer:
reduced expression of PPARg-coactivator1 (PGC-1). Cancer Lett. 203:25–33.
139. Zhang Y, Ba Y, Liu C, Sun G, Ding L, Gao S and Hao J, et al. 2007, PGC-1a induces
apoptosis in human epithelial ovarian cancer cells through a PPARg-dependent pathway. Cell
Res. 17:363–373.
140. Cui L, Jeong H, Borovecki F, Parkhurst CN, Tanese N, Krainc D. 2006, Transcriptional
repression of PGC-1α by mutant huntingtin leads to mitochondrial dysfunction and
neurodegeneration. Cell. 127:59–69.
141. Vianna CR, Huntgeburth M, Coppari R, Choi CS, Lin J, Krauss S, et al. 2006,
Hypomorphic mutation of PGC-1β causes mitochondrial dysfunction and liver insulin resistance.
Cell Metab. 4:453–464.
142. Kadonaga JT, Tjian R. 1986, Affinity purification of sequence-specific DNA binding
proteins. PNAS. 83:5889-5893.
92
143. Yang TTC, Chow CW. 2012, Elucidating protein:DNA complex by oligonucleotide DNA
affinity purification. Methods Mol Biol. 809:75-84.
144. Sleeman KE, Kendrick H, Robertson D, Isacke CM, Ashworth A, Smalley MJ. 2007,
Dissociation of estrogen receptor expression and in vivo stem cell activity in the mammary gland.
J Cell Biol. 176:19-26.
145. Sleeman KE, Kendrick H, Ashworth A, Isacke CM, Smalley MJ. 2006, CD24 staining of
mouse mammary gland cells defines luminal epithelial, myoepithelial/basal and non-epithelial
cells. Breast Cancer Res. 8:R7.
146. Holstege H, Horlings HM, Velds A, Langerød A, Børresen-Dale AL, van de Vijver MJ,
Nederlof PM, Jonkers J. 2010, BRCA1-mutated and basal-like breast cancers have similar
aCGH profiles and a high incidence of protein truncating TP53 mutations. BMC Cancer. 10:654.
147. Hitchler MJ, Domann FE. 2, 2009, Metabolic defects provide a spark for the epigenetic
switch in cancer. Free Radic Biol Med. 47:115-127.
148. Stampfer MR, Garbe J, Levine G, Lichtsteiner S, Vasserot AP, Yaswen P. 2001,
Expression of the telomerase catalytic subuint, hTERT, induces resistance to transforming growth
factor beta growth inhibition in p16INK4A(-) human mammary epithelial cells. PNAS. 98:44984503.
149. Debnath J, Muthuswamy SK, Brugge JS. 2003, Morphogenesis and oncogenesis of MCF10A mammary epithelial acini grown in three-dimensional basement membrane cultures.
Methods. 30:256-268.
150. Weaver VM, Bissell MJ. 1999, Functional culture models to study mechanisms governing
apoptosis in normal and malignant mammary epithelial cells. J Mammary Gland Biol Neoplasia.
4:193-201.
151. Debnath J, Mills KR, Collins NL, Reginato MJ, Muthuswamy SK, Brugge JS. 2002,
The role of apoptosis in creating and maintaining luminal space within normal and oncogeneexpressing mammary acini. Cell. 111:29-40.
152. Fromm L, Burden SJ. 2001, Neuregulin-1-stimulated phosphorylation of GABP in skeletal
muscle cells. Biochemistry. 40:5306-5312.
153. Gohil VM, Sheth SA, Nilsson R, Wojtovich AP, Lee JH, Perocchi F, Chen W, Clish CB,
Ayata C, Brookes PS, Mootha VK. 2010, Nutrient-sensitized screening for drugs that shift
energy metabolism from mitochondrial respiration to glycolysis. Nature Biotechnol. 28:249-55.
93
154. Marroquin LD, Hynes J, Dykens JA, Jamieson JD, Will Y. 2007, Circumventing the
Crabtree effect: Replacing media glucose with galactose increases susceptibility of HepG2 cells
to mitochondrial toxicants. Toxicol Sci. 97:539-547.
155. Piantadosi CA, Suliman HB. 2006, Mitochondrial transcription factor A induction by
redox activation of nuclear respiratory factor 1. J Biol Chem. 281:324-333.
156. Kluck RM, Bossy-Wetzel E, Green DR, and Newmeyer DD. 1997, The release of
cytochrome c from mitochondria: a primary site for Bcl-2 regulation of apoptosis. Science.
275:1132-1136.
157. Liu X, Kim CN, Yang J, Jemmerson R, and Wang X. 1996, Induction of apoptotic
program in cell-free extracts: requirement for dATP and cytochrome c. Cell. 86:147-157.
158. Evans MJ, Scarpulla RC. 1989 , Interaction of nuclear factors with multiple sites in the
somatic cytochrome c promoter. Characterization of upstream NRF-1, ATF, and intron Sp1
recognition sequences. J Biol Chem. 264:14361-14368.
159. Vempati UD, Han X, Moraes CT. 2009, Lack of cytochrome c in mouse fibroblasts
disrupts assembly/stability of respiratory complexes I and IV. J Biol Chem. 284:4383-4391.
160. Nedvetskya PI, Kwona SH, Debnath J, Mostov KE. 2012, Cyclic AMP regulates
formation of mammary epithelial acini in vitro. Mol Biol Cell. 23:2973-2981 .
161. . Herzig S, Long F, Jhala US, Hedrick S, Quinn R, Bauer A, et al. 2001, CREB regulates
hepatic gluconeogenesis through the coactivator PGC-1. Nature. 413:179-183.
162. Guo J, Zheng L, Liu W, Wang X, Wang Z, Wang Z, et al. 2011, Frequent truncating
mutation of TFAM induces mitochondrial DNA depletion and apoptotic resistance in
microsatellite-unstable colorectal cancer. Cancer Res. 71:2978 .
163. Dey R, Moraes CT. 2000, Lack of oxidative phosphorylation and low mitochondrial
membrane potential decrease susceptibility to apoptosis and do not modulate the protective effect
of Bcl-xL in osteosarcoma cells. J Biol Chem. 275:7087-7094.
164. Li K, Li Y, Shelton JM, Richardson JA, Spencer E, Chen ZJ, Wang X, Williams RS.
2000, Cytochrome c deficiency causes embryonic lethality and attenuates stress-induced
apoptosis. Cell. 101:389–399.
165. Kwong JQ, Henning MS, Starkov AA, Manfredi G. 2007, The mitochondrial respiratory
chain is a modulator of apoptosis. J Cell Biol. 17:1163–1177.
94
166. Humphreys RC, Krajewska M, Krnacik S, Jæger R, Weiher H, Krajewski S, Reed JC,
Rosen JM. 1996, Apoptosis in the terminal endbud of the murine mammary gland: a mechanism
of ductal morphogenesis. Development. 122:4013-4022.
167. Cam H, Balciunaite E, Blais A, Spektor A, Scarpulla RC, Young R, Kluger Y, Dynlacht
BD. 2004, A common set of gene regulatory networks links metabolism and growth inhibition.
Mol Cell. 16:399–411.
168. Ma Y, Bai RK, Trieu R, Wong LJ. 2010, Mitochondrial dysfunction in human breast
cancer cells and their transmitochondrial cybrids. Biochim Biophys Acta. 1797:29-37.
169. Springer EL. 1980, Comparative study of the cytoplasmic organelles of epithelial cell lines
derived from human carcinomas and nonmalignant tissues. Cancer Res. 40:803-817.
170. Simonnet H, Alazard N, Pfeiffer K, Gallou C, Béroud C, Demont J, Bouvier R,
Schägger H, Godinot C. 2002, Low mitochondrial respiratory chain content correlates with
tumor aggressiveness in renal cell carcinoma. Carcinogenesis. 23:759-768.
171. Shapovalov Y, Hoffman D, Zuch D, de Mesy Bentley KL, Eliseev RA. 2011,
Mitochondrial dysfunction in cancer cells due to aberrant mitochondrial replication. J Biol Chem.
286:22331-22338.
172. Bai RK, Chang J, Yeh KT, Lou MA, Lu JF, Tan DJ, Liu H, Wong LJC. 2011,
Mitochondrial DNA content varies with pathological characteristics of breast cancer. J Oncol.
10.1155/2011/496189
173. Michelakis ED, Webster L, Mackey JR. 2008, Dichloroacetate (DCA) as a potential
metabolic-targeting therapy for cancer. Br J Cancer. 99:989-994.
174. Prigione A, Adjaye J. 2010, Modulation of mitochondrial biogenesis and biogenergetic
metabolism upon in vitro and in vivo differentiation of human ES and iPS cells. Int J Dev Biol.
54:1729-1741.
175. Juntilla MM, Patil VD, Calamito M, Joshi RP, Birnbaum MJ, Koretzky GA. 2010,
AKT1 and AKT2 maintain hematopoietic stem cell function by regulating reactive oxygen
species. Blood. 115:4030-4038 .
176. Tormos KV, Anso E, Hamanaka RB, Eisenbart J, Joseph J, Kalyanaraman B, Chandel
NS. 2011, Mitochondrial complex III ROS regulate adipocyte differentiation. Cell Metab.
14:537-544.
95
177. Gambhir SS, Czernin J, Schwimmer J, Silverman DHS, Coleman RE, Phelps ME.
2001, A tabulated summary of the FDG PET literature. J Nucl Med. 42:suppl 1S-93S .
178. Kroemer G, Pouyssegur J. 2008, Tumor Cell Metabolism: Cancer's Achilles' Heel. Cancer
Cell. 13:472-482.
179. Helmlinger G, Yuan F, Dellian M, Jain RK. 1997, Interstitial pH and pO2 gradients in
solid tumours in vivo: high resolution measurements reveal a lack of correlation. Nature Med.
3:177-182.
180. Patel AA, Gawlinski ET, Lemieux SK, Gatenby RA. 2001, A cellular automaton model of
early tumor growth and invasion: the effects of native tissue vascularity and increased anaerobic
tumor metabolism. J Theor Biol. 213:315-331.
181. Fischer K, Hoffmann P, Voelkl S, Meidenbauer N, Ammer J, Edinger M, et al. 2007,
Inhibitory effect of tumor cell-derived lactic acid on human T cells. Blood. 109:3812-3819.
182. Koukourakis MI, Giatromanolaki A, Harris AL, Sivridis E. 2006, Comparison of
metabolic pathways between cancer cells and stromal cells in colorectal carcinomas: a metabolic
survival role for tumor-associated stroma. Cancer Res. 66:632-637.
183. Johnson RF, Perkins ND. 2012, Nuclear factor-kB, p53, and mitochondria: regulation of
cellular metabolism and the Warburg effect. TIBS. 37:317-324.
184. Zhou W, Choi M, Margineantu D, Margaretha L, Hesson J, Cavanaug C. 2012, HIF1α
induced switch from bivalent to exclusively glycolytic metabolism during ESC-to-EpiSC/hESC
transition. EMBO. 31:2103-2116 .
185. Dhar SS, Ongwijitwat S, Wong-Riley MT. 2008, Nuclear respiratory factor 1 regulates all
ten nuclear-encoded subunits of cytochrome c oxidase in neurons. J Biol Chem. 283:3120-3129.
186. Seagroves TN, Ryan HE, Lu H, Wouters BG, Knapp M, Thibault P, Laderoute K,
Johnson RS. 2001, Transcription factor HIF-1 is a necessary mediator of the pasteur effect in
mammalian cells. Mol Cell Biol. 21:3436-3444.
187. Zhang H, Gao P, Fukuda R, Kumar G, Krishnamachary B, Zeller KI, Dang CV,
Semenza GL. 2007, HIF-1 inhibits mitochondrial biogenesis and cellular respiration in VHLdeficient renal cell carcinoma by repression of C-MYC activity. Cancer Cell. 11:407-420.
188. Seagroves TN, Hadsell D, McManaman J, Palmer C, Liao D, McNulty W, et al. 2003,
HIF1alpha is a critical regulator of secretory differentiation and activation, but not vascular
expansion, in the mouse mammary gland. Development. 130:1713-1724.
96
189. Seyfried TN, Shelton LM. 2010, Cancer as a metabolic disease. Nutr Metab. 7:1-22.
190. Suliman HB, Sweeney TE, Withers CM, Piantadosi CA. 2010, Co-regulation of nuclear
respiratory factor-1 by NFkappaB and CREB links LPS-induced inflammation to mitochondrial
biogenesis. J Cell Science. 23:2565-2575.
191. Kenneth NS, Rocha S. 2008, Regulation of gene expression by hypoxia. J Biochem.
414:19-29.
192. Compton S, Kim C, Griner NB, Potluri P, Scheffler IE, Sen S, et al. 2011, Mitochondrial
dysfunction impairs tumor suppressor p53 expression/function. J Biol Chem. 286:20297-20312.
97