Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Microenvironmental and biomechanical regulation of mitochondrial membrane potential in cancer cells
(USC Thesis Other)
Microenvironmental and biomechanical regulation of mitochondrial membrane potential in cancer cells
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
MICROENVIRONMENTAL AND BIOMECHANICAL REGULATION OF
MITOCHONDRIAL MEMBRANE POTENTIAL IN CANCER CELLS
by
Hydari Masuma Begum
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(BIOMEDICAL ENGINEERING)
May 2022
Copyright 2022 Hydari Masuma Begum
ii
Acknowledgements
I would like to extend my heartfelt gratitude to the USC Graduate School for giving me the
opportunity to pursue my Ph.D. I am very grateful for my mentor, Dr. Keyue Shen, who has
painstakingly and with great patience taught me the fundamentals of research from cell culture to
compiling manuscripts. I admire and respect his work and his dedication to science and will always
be grateful for all the experiences I had during my time in Shen Lab. I am also very grateful for
my PhD defense committee Dr. Michael Khoo, Dr. Min Yu, Dr. Stacey Finley and Dr. Shannon
Mumenthaler for their valuable feedback and guidance during my qualifying examination.
I would not have looked forward to spending so much time in the lab had it not been for
the amazing people I was surrounded by. Dr. Yuta Ando, Dr. Gunce Cinay, thank you for letting
me tag along during your very first semester of the Ph.D. and patiently explaining your research
and experimental techniques to me. Yuta, thank you for being an amazing labmate and friend, my
time in the lab would not have been the same without the discussions we used to have about
research and otherwise, all the VGSA events we did together, and the mid-day snack runs. Hao
Zhou, Jia Hao, Jeong Min Oh, thank you for being such great labmates. I always enjoyed our
discussions and chats in the lab office, and I know my time here wouldn’t have been the same
without you all. Hao and Jia, I am so excited and honored to be graduating with you both this year.
Shoutout to Jeong for taking charge of lab safety inspections to ensure everything runs smoothly,
I am sure that you will be an amazing senior Ph.D. student mentor to the next batch of Shen Lab
students. I would also like to thank Chelsea Mariano, Kristen Nemes and Irene Li, for helping me
with my research and all the lively discussions we had during journal clubs and lab meetings.
I would like to thank my family for always believing in me. Papa, Mom, Ali, I am forever
indebted to you for your love, kindness and support beyond measure. This arduous journey would
iii
not have been possible had it not been for you all and I feel incredibly grateful to have your love
and support through all of this. Lastly, I would like to thank my friends who became family during
my time here in a new country and city, thousands of miles away from home. Dr. Zahoor Sadiq,
Dr. Ali Marjaninejad, Dr. Zumra Seidel, Dr. Scott Seidel, Celia Brown, Dr. Brian Cohn, Srishti
Saigal, Thanmayi Bhasvaraj, I am so grateful for your friendship and support during these years.
iv
Table of Contents
Acknowledgements ..................................................................................................................................... ii
List of tables............................................................................................................................................... vii
List of figures ............................................................................................................................................ viii
Abstract......................................................................................................................................................... x
Chapter 1: Introduction .............................................................................................................................. 1
1.1 How is the ΔΨm established? ................................................................................................ 3
1.2 ΔΨm of normal vs. cancer cells ............................................................................................. 5
1.3 Functional significance of the ΔΨm ....................................................................................... 6
1.3.1 ATP synthesis: ................................................................................................................ 6
1.3.2 Reactive oxygen species (ROS) production ................................................................... 7
1.3.2 Mitochondrial quality control ......................................................................................... 8
1.3.3: Mitochondrial and cellular motility ............................................................................. 11
1.4 Intrinsic regulation of the ΔΨm............................................................................................ 11
1.4.1 Regulation of ΔΨm by inner membrane ion transporters............................................. 13
1.4.2 Regulation of ΔΨm by outer membrane ion transporters............................................. 15
1.4.3 ΔΨm regulation by cytoskeletal elements .................................................................... 15
1.4.4 Regulation of ΔΨm by signaling pathways .................................................................. 19
1.5 How does the TME regulate ΔΨm of cancer cells? ............................................................. 19
1.5.1: Hypoxia ....................................................................................................................... 20
1.5.2 Substrate Stiffness and Composition ............................................................................ 21
1.5.3 Compressive and tensile stresses .................................................................................. 21
1.6 Techniques to investigate ΔΨm ........................................................................................... 22
1.6.1 In vivo techniques......................................................................................................... 22
1.6.2 In vitro techniques ........................................................................................................ 23
1.7 Objectives and aims ............................................................................................................ 23
1.8 Novelty and Impact ............................................................................................................. 26
Chapter 2: Spatial Regulation of Mitochondrial Heterogeneity by Stromal Confinement in
Micropatterned Tumor Models ............................................................................................................... 27
v
2.1 Rationale.............................................................................................................................. 27
2.2 Materials and Methods ........................................................................................................ 29
2.2.1 Cell Culture .................................................................................................................. 29
2.2.2 Antibodies..................................................................................................................... 29
2.2.3 Micropatterned Tumor-Stromal Assay (μTSA) ........................................................... 29
2.2.4 Imaging of the Mitochondrial Membrane Potential (ΔΨm) in the μTSA ..................... 30
2.2.5 Immunofluorescence .................................................................................................... 30
2.2.6 Metabolic Imaging (NAD(P)H and FAD fluorescence) .............................................. 31
2.2.7 Laser Capture Microdissection (LCM) and RNA Sequencing..................................... 31
2.2.8 Gene Set Enrichment Analysis (GSEA) ....................................................................... 32
2.2.9 Mechanical Constraints ................................................................................................ 33
2.2.10 In vivo Metastasis Model ........................................................................................... 34
2.2.11 Statistical Analysis ..................................................................................................... 34
2.3 Results ................................................................................................................................. 35
2.3.1 Mitochondrial and metastatic pathways are upregulated at the tumor-stromal interface
in the µTSA ........................................................................................................................... 35
2.3.2 Mitochondrial membrane potential (ΔΨm) is spatially regulated in the µTSA ............ 37
2.3.3 Fluorescence microscopy of metabolic coenzymes reveals enhanced redox activity at
the interface ........................................................................................................................... 39
2.3.4 ΔΨm spatial profile correlates with YAP/TAZ nuclear localization in the micropatterns
............................................................................................................................................... 41
2.3.5 ΔΨm distribution and YAP/TAZ nuclear translocation in µTSA are modulated by
stromal density....................................................................................................................... 44
2.3.6 Inhibiting the Rho-associated protein kinase and actin polymerization leads to a loss of
ΔΨm spatial distribution ......................................................................................................... 48
2.3.7 ΔΨm level correlates with metastatic potential in vivo ................................................. 51
2.4 Discussion ........................................................................................................................... 54
Chapter 3: E-Cadherin Regulates Mitochondrial Membrane Potential in Cancer Cells ................. 60
3.1 Rationale.............................................................................................................................. 60
3.2 Materials and Methods ........................................................................................................ 62
3.2.1 Cell culture and micropatterning .................................................................................. 62
3.2.2 Generation of E-cadherin-GFP expressing and E-cadherin knockout cell lines .......... 63
3.2.3 Mitochondrial membrane potential staining and imaging ............................................ 64
3.2.4 Drug treatment and immunostaining ............................................................................ 64
3.2.5 Image analysis and quantification ................................................................................ 65
vi
3.2.6 Statistical analysis ........................................................................................................ 65
3.3 Results ................................................................................................................................. 66
3.3.1 Adherens junctions (AJs) are downregulated at the tumor-stromal interface in a
micropatterned tumor model ................................................................................................. 66
3.3.2 E-cadherin expression correlates with spatial distribution of m within tumor
micropattern ........................................................................................................................... 68
3.3.3 Disrupting AJ formation increases m in MCF-7 micropattern ................................ 71
3.3.4 E-cadherin expression in MDA-MB-231 cells decreases ΔΨm at the micropattern
center ..................................................................................................................................... 74
3.3.5 E-cadherin knockout and overexpression alter m at the center of MCF-7
micropattern ........................................................................................................................... 78
3.4 Discussion ........................................................................................................................... 81
Chapter 4: Regulation of Mitochondrial Membrane Potential by YAP in Cancer Cells ................ 83
4.1 Rationale.............................................................................................................................. 83
4.2 Materials and Methods ........................................................................................................ 85
4.2.1 Cell Culture .................................................................................................................. 85
4.2.2 Generation of EGFP-YAP Expressing and YAP Knockdown Cell Lines ................... 85
4.2.3 Assessing mitochondrial membrane potential using flow cytometry........................... 86
4.2.4 Live-tracking mitochondrial membrane potential and EGFP-YAP ............................. 87
4.2.5 Mitochondrial fractionation .......................................................................................... 87
4.2.6 Western Blotting ........................................................................................................... 88
4.2.7 Gene expression assays (qPCR) ................................................................................... 89
4.3 Results ................................................................................................................................. 89
4.3.1 YAP overexpression decreases the m of MCF-7 breast cancer cells ....................... 89
4.3.2 YAP inhibition by verteporfin decreases m whereas YAP activation by TRULI
increases the m of MCF-7 cells ........................................................................................ 92
4.3.3 Depolarization of the m does not alter EGFP-YAP fluorescence intensity and
localization ............................................................................................................................ 95
4.3.4 YAP can be detected in the mitochondrial protein isolates of MCF-7 cells ................ 98
4.3.5 YAP k/d leads to an increase in mitochondrial transcription of ETC genes .............. 101
4.3.6 YAP overexpression decreases mitochondrial respiration and increases glycolysis in
MCF-7 cells ......................................................................................................................... 104
4.4 Discussion ......................................................................................................................... 107
References ................................................................................................................................................ 114
vii
List of tables
Table 4-1: List of TaqMan assays used for qPCR
viii
List of figures
Figure 1-1: Schematic showing the tricarboxylic acid (TCA) cycle and electron transport chain
(ETC) within the mitochondria.
Figure 1-2: Schematic showing intrinsic (top) and extrinsic (bottom) factors that could regulate
cancer cell ΔΨm.
Figure 2-1: Spatial regulation of mitochondrial, metabolic, and metastatic pathways in
micropatterned tumor-stromal assays (µTSA) assessed by RNA sequencing.
Figure 2-2: Differential regulation of mitochondrial membrane potential (ΔΨm) and mass at the
interface vs. center in the µTSA.
Figure 2-3: Redox imaging of cancer cells in a µTSA.
Figure 2-4: Correlation of ΔΨm and YAP/TAZ nuclear translocation in micropatterns.
Figure 2-5: Regulation of cancer cell ΔΨm by stromal density.
Figure 2-6: Loss of ΔΨm heterogeneity by inhibition of mechanotransduction.
Figure 2-7: Correlation of ΔΨm with metastatic potential in vivo.
Figure 2-S1: Fluorescence images showing F-actin staining in MCF-7 cells at the centers and
edges of day 4 micro-patterns, after 4 hours of drug treatment with 50μM Y-27632 or 0.5μM
Latrunculin A, along with their respective no treatment controls.
Figure 2-S2: Quantification of nuclear YAP/TAZ in MCF-7 cells at the centers and edges of day
4 micropatterns, following 12 hours of drug treatment with 50μM Y-27632 or 0.5μM Latrunculin
A, along with their respective no treatment controls.
Figure 3-1: Spatial distribution of m of MCF-7 cells in micropatterned tumor model
associated with regulation of cell adhesion.
ix
Figure 3-2: Correlation of E-cadherin expression with spatial distribution of m.
Figure 3-3: Disruption of AJs with DTT in MCF-7 micropatterns.
Figure 3-4: Effect of E-cadherin expression on the spatial distribution of ΔΨm in MDA-MB-231
micropatterns.
Figure 3-5: Effects of E-cadherin knockout (KO) and overexpression (OE) on m in MCF-7
micropatterns.
Figure 4-1: Effect of YAP overexpression (OE) on the m of MCF-7 cells.
Figure 4-2: Effect of YAP inhibition (verteporfin) and activation (TRULI) on the m of MCF-7
cells.
Figure 4-3: Effect of FCCP-induced m depolarization on cytoplasmic and nuclear YAP
fluorescence.
Figure 4-4: Identification of YAP in mitochondrial protein of MCF-7 cells.
Figure 4-5: Effect of YAP k/d on mitochondrial transcription.
Figure 4-6: Effect of YAP OE on mitochondrial bioenergetics of MCF-7 cells.
Figure 4S-1: Effect of YAP OE on glycolytic metabolism of MCF-7 cells.
Figure 4S-2: Effect of drug treatments on FITC autofluorescence of MCF-7 cells.
x
Abstract
Cancer cells have an abnormally high mitochondrial membrane potential (ΔΨm), which is
associated with enhanced invasive properties in vitro. The mechanisms underlying the abnormal
ΔΨm in cancer cells remain unclear. Research on different cell types has shown that ΔΨm is
regulated by various intracellular mechanisms such as by mitochondrial inner and outer membrane
ion transporters, cytoskeletal elements and biochemical signaling pathways. On the other hand, the
role of extrinsic, tumor microenvironment (TME) derived cues in regulating ΔΨm is not well
defined. In this dissertation, we first summarize the existing literature on intercellular mechanisms
of ΔΨm regulation, with a focus on cancer cells. We then offer our perspective on the different
ways through which the microenvironmental cues such as mechanical stresses may regulate cancer
cell ΔΨm. In the subsequent chapters we present our work on recapitulating mitochondrial
heterogeneity in vitro using a micropatterning based approach and the mechanisms that regulate
the spatial gradients of ΔΨm observed in them.
1
Chapter 1: Introduction
Nearly a century ago, Otto Warburg discovered that cancer cells have enhanced levels of
glycolysis, and therefore abnormally high glucose uptake and lactate production even in the
absence of hypoxic conditions (a phenomenon he called ‘aerobic glycolysis’)
1
. It is now known
that this phenomenon, i.e., the Warburg effect, is not a universal characteristic of all cancer cell
types. Rather, tumors are highly heterogenous comprising of subpopulations of cancer cells
displaying ‘Warburg-like’ metabolic characteristics
2
. Nevertheless, this increased propensity for
glucose consumption by some cancer cell subpopulations is routinely exploited for diagnostic
fluorodeoxyglucose-positron emission tomography (FDG-PET) scans in many cancer types
including breast and lung cancers
3
. Furthermore, such metabolic heterogeneity has emerged as an
important cancer cell hallmark and a potential therapeutic target
4,5
. Interestingly, studies
investigating the mitochondrial activity of these metabolically heterogenous cancer cells have
found an association between dysregulated mitochondrial activity and their invasiveness and
metastatic potential
6-9
. A mechanistic understanding of the factors that lead to altered cancer cell
mitochondrial activity is thus crucial to develop novel therapeutics, for instance, by targeting the
mitochondrial support of metastatic behaviors.
A unique observation of cancer cell mitochondrial function is that many epithelial cancer
cell types have an unusually high mitochondrial membrane potential ( m), compared to their
normal counterparts
10,11
. Further studies showed an association of cancer cell m with a range of
cell behaviors including decreased susceptibility to apoptosis
12
and the acquisition of an invasive
phenotype
13
. The ΔΨm is thus associated with the development of important cancer cell
hallmarks
14
, and a greater understanding of the mechanisms by which it affects cell phenotypes
2
will help in the discovery of novel therapeutic targets. However, whether the high ΔΨm is related
to the ‘Warburg-phenotype’, and how it is differentially regulated in cancer cells, remains unclear.
One possibility is that the abnormally high m in cancer cells is a result of genetic
mutations that affect their intrinsic mitochondrial activity. Indeed, mutations in mitochondrial
DNA (mtDNA) and variations in its copy number have been reported in several types of
cancers
5,15
. Furthermore, it is possible that the altered m results from the activity of oncogenes
such as MYC, AKT, RAS, BRAF and reduction of tumor suppressors like p53, which were found
to affect various metabolic pathways including those involving mitochondrial metabolism
4,16,17
.
While it is possible that the altered cancer cell m is primarily a result of these cancer-specific
mutations, there is also growing evidence supporting alternative mechanisms of physical
interaction and biochemical signaling taking place within the tumor microenvironment (TME)
18
.
In this chapter, we will briefly summarize how the m is established and regulated, with a focus
on TME-derived cues that have the capacity to regulate cancer cell m. We will then discuss the
current in vitro and in vivo techniques used to probe m, and their potential applications in
investigating m-driven tumor progression.
3
1.1 How is the ΔΨm established?
Figure 1-1: Schematic showing the tricarboxylic acid (TCA) cycle and electron transport chain
(ETC) within the mitochondria.
Mitochondria are double membrane bound organelles. Transport of metabolites and ions
across the outer mitochondrial membrane is regulated by large porin protein channels such as
Voltage-Dependent Anion Channels (VDACs) that are permeable to solutes up to 5 kDa
19
. On the
other hand, the mitochondrial inner membrane is largely impermeable, and transport of molecules
across this membrane requires specialized membrane transport proteins
19,20
. The inner membrane
also forms deep invaginations into the mitochondrial matrix, called cristae (Fig. 1). The presence
of cristae on the inner mitochondrial membrane greatly increases its effective surface area. It is
also lined with respiratory complexes of the Electron Transport Chain (ETC) and is the site of
4
mitochondrial adenosine triphosphate (ATP) synthesis. The mitochondrial matrix contains various
enzymes that support the replication and transcription of mtDNA
20
. In addition, it also the site of
the tricarboxylic acid (TCA) cycle and contains TCA cycle associated enzymes.
Tricarboxylic acid (TCA) cycle: Pyruvate from the cytosol (one of the end products of
glycolysis) is transported into the mitochondrial matrix through a pyruvate carrier present on the
inner mitochondrial membrane and converted to acetyl coenzyme A (acetyl CoA) by the enzyme
pyruvate dehydrogenase. Acetyl CoA enters the TCA cycle, where it is oxidized to produce carbon
dioxide and water
21
. During this process, electron carrier molecules nicotinamide adenine
dinucleotide (NAD) and flavine adenine dinucleotide (FAD) are reduced, forming NADH and
FADH2 from NAD
+
and FAD respectively. These reduced electron carriers then take part in the
ETC in the inner mitochondrial membrane
21
(Fig. 1).
Electron transport chain (ETC): The ETC takes place in the inner mitochondrial
membrane and oxidizes electron carriers from the TCA cycle. Briefly, electrons from NADH and
FADH2 are transferred to coenzyme Q by complexes I and II of the ETC respectively. Coenzyme
Q carries these electrons to complex III, which donates them to cytochrome c (cyt c). Cyt c then
diffuses along the membrane and transfers the electrons to complex IV, which then transfers these
electrons to oxygen, eventually producing water
21
. In complexes I, III and IV, the activity of
electron transfer is coupled to the pumping of protons into the mitochondrial intermembrane
space
21
(Fig. 1-1). This generates an electrochemical proton gradient across the mitochondrial
inner membrane, which gives rise to the mitochondrial membrane potential (ΔΨm).
5
1.2 ΔΨm of normal vs. cancer cells
More than thirty years ago, Summerhayes et al., first reported that most in vitro cultures of
epithelial cancer cells have a much higher ΔΨm (as assessed by uptake and retention of ΔΨm-
sensitive dye rhodamine 123) when compared to non-transformed cells
10
. Apart from these cancer
cells, they found cardiac muscle cells and myotubes to possess such “unusually” high levels of
ΔΨm
10
. Other cancer cell types screened from blood cancers, connective tissue derived cancers
(osteosarcoma) and neuroblastomas do not have the high ΔΨm that is characteristic of epithelial
carcinomas
11
. To find out whether intrinsic differences in the ΔΨm of cancer cells can affect tumor
growth and progression, Heerdt et al., subcloned isogenic colonic carcinoma cells with stable
differences in their ΔΨm
12
. They found that the intrinsic ΔΨm differences of these cells are
associated with differences in mitochondrial mRNA levels, butyrate-induced cell cycle arrest and
susceptibility to apoptosis
12
. Further studies with these cells showed that the high-ΔΨm subclones
have higher vascular endothelial growth factor (VEGF) and matrix metalloproteinase 7 (MMP7)
secretion, as well as higher invasiveness
13,22
. Similarly, high-ΔΨm breast cancer (MCF-7)
subclones are also shown to have greater VEGF secretion than the low-ΔΨm subclones, suggesting
a correlation of ΔΨm with the invasive properties in these cells
23
.
Bonnet et al., attempted to selectively target the high-ΔΨm of cancer cells. They found that
treatment with dichloroacetate (DCA), a pharmacological inhibitor of mitochondrial pyruvate
dehydrogenase kinase (PDK), selectively lowers the ΔΨm of cancer cells (breast and lung
carcinoma cells, as well as glioma cells) while having no effect on the ΔΨm of non-cancerous
epithelial cells
24
. Further, DCA treatment also causes release of cytochrome c from the
mitochondria, thus inducing apoptosis
24
. In in vivo studies, it was found that DCA treatment can
reduce tumor cell proliferation and increase their apoptosis, thus reducing the overall tumor size
24
.
6
Most studies that reported observations of high ΔΨm of cancer cells were performed using
in vitro cultures. Recently, Momcilovic et al., used a positron emission tomography (PET) imaging
technique to investigate the ΔΨm of lung cancer cells in vivo
25
. They found a tumor-subtype
dependent ΔΨm heterogeneity in these mouse models – adenocarcinomatous tumors have a higher
ΔΨm whereas small squamous cell carcinoma tumors have lower ΔΨm
25
. In summary, cancer cells
are found to have highly dysregulated mitochondrial functions
26
, which can manifest as differences
in their ΔΨm
11
.
1.3 Functional significance of the ΔΨm
1.3.1 ATP synthesis: Complex V of the ETC is an ATP synthase. Driven by the proton
gradient generated by the ETC, it pumps protons back into the mitochondrial matrix and uses this
energy to combine a molecule of adenosine diphosphate (ADP) with an inorganic phosphate to
produce ATP
21
. This process of ATP production is termed oxidative phosphorylation (OXPHOS)
since it requires the presence of oxygen. The ATP produced by the mitochondrial ATP synthase
is transported out of the mitochondrial matrix in exchange for a molecule of cytosolic ADP by an
adenine nucleotide transporter (ANT) (Fig. 1). Under some conditions of stress (such as ischemia),
it has been reported that the ATP synthase and ANT reverse their activity – the ANT now
transports ATP into the mitochondrial matrix in exchange for ADP, and the ATP synthase utilizes
this ATP to form ADP and pump out protons into the mitochondrial intermembrane space, in an
attempt to maintain the ΔΨm
27
. In this state, mitochondria become major consumers of leftover
cellular ATP, and try to maintain the ΔΨm in order to delay the onset of autophagic or apoptotic
processes (which are associated with sustained ΔΨm depolarization). Zorova et al., suggest that the
7
maintenance of ΔΨm at the expense of cellular ATP levels demonstrates how important it is for the
cells to maintain stable ΔΨm levels
28,29
.
1.3.2 Reactive oxygen species (ROS) production: As explained in the previous section,
there is a transport or shuttling of electrons across the complexes of the ETC in the inner
mitochondrial membrane. However, this electron transfer is not perfect, and there is electron
leakage at Complex I (in the direction of the mitochondrial matrix) and at Complex III (in both
directions of the mitochondrial matrix and the intermembrane space)
30
. This leads to the partial
reduction of molecular oxygen (O2). Partially reduced forms of O2 are highly reactive free radicals.
These include the superoxide anion and hydrogen peroxide and together with hydroxyl ions, are
called Reactive Oxygen Species (ROS)
31
. In isolated rat heart mitochondria, ROS levels are low
and unaffected by values of ΔΨm below 140 mV
32,33
. Any increases in the ΔΨm beyond that
threshold value leads to very high increases in ROS levels
32
. Inhibiting the ATP synthase using
oligomycin in cancer cells led to ΔΨm hyperpolarization along with increased ROS levels
34
. These
results suggest that a higher ΔΨm is associated with high ROS levels. However, in pathological
conditions that affect the ETC complexes, low-ΔΨm levels have also been associated with high
ROS levels
35
. Similarly, antimycin A (ETC Complex III inhibitor) treatment leads to a decrease in
ΔΨm levels and increased mitochondrial ROS production
34
. In summary, in mitochondrial
disorders related to mitochondrial ATP synthase, increased ΔΨm is associated with high ROS
levels; in disorders related to ETC subunits, decreased ΔΨm is associated with high ROS levels
34
.
To explain this apparent paradox, Li et al., propose an “ROS balance” hypothesis, suggesting that
there is an optimal level of ΔΨm that is associated with physiological (low) levels of ROS
30
. Any
extreme fluctuations in the ΔΨm levels (too low or too high) would lead to high ROS levels either
8
by affecting the rate of electron transfer through the ETC or disrupting the balance of antioxidant
enzymes
30
.
1.3.2 Mitochondrial quality control: The mitochondrial network within cells is not static.
It consists of motile mitochondria
36
that can fuse together when they encounter each other (also
called mitochondrial fusion) or split into daughter mitochondria (a process called mitochondrial
fission). Together with mitophagy (selective elimination of damaged mitochondria), these
processes form the cells’ quality control strategy for maintaining the bioenergetic efficiency and
integrity of their mitochondrial networks
37
. The role of ΔΨm in regulating the processes of
mitochondrial fusion, fission and mitophagy is examined below.
Mitochondrial fusion: The process of mitochondrial fusion facilitates rapid diffusion of
mitochondrial content, allowing exchange of nutrients and metabolites within the mitochondrial
network
38
. Long-term tracking of fusion events in individual mitochondria has led to the finding
that fusion occurs as frequently as one event every 5-20 minutes per mitochondrion in cell lines
derived from rodent kidney and pancreatic beta-cells
37,39
. Mitofusins Mfn1 and Mfn2 mediate
mitochondrial outer membrane fusion whereas Opa1 mediates fusion of inner mitochondrial
membranes
40
. Fusion appears to be a selective event based on ΔΨm levels: mitochondria with high-
ΔΨm have a greater probability to undergo fusion when compared to low-ΔΨm mitochondria
39
;
reduction of ΔΨm by uncoupling drugs hinders fusion and enhances fragmentation of the
mitochondrial network
39,41,42
. Depolarizing, i.e., a reduction in the ΔΨm is associated with
proteolytic processing and therefore inactivation of OPA-1
43,44
, and this could be a potential
mechanism by which the ΔΨm regulates fusion. ΔΨm repolarization was found to restore fusion in
breast cancer cells
42
. Together, these studies suggest that the process of mitochondrial fusion
9
requires an intact ΔΨm, and controlling the ΔΨm can regulate mitochondrial fusion events.
Conversely, inhibiting the mitofusin proteins leads to a decrease in the overall ΔΨm of the cells,
suggesting that the overall ΔΨm of the cells is maintained through fusion by allowing intermixing
of mitochondrial matrix as well as membrane components
38
.
Mitochondrial fission: Mitochondrial fission is the process by which an individual
mitochondrion splits into two daughter mitochondria. Interestingly, this form of mitochondrial
division appears to be asymmetrical – more than four-fifths of all tracked fission events led to the
formation of a depolarized and a hyperpolarized daughter mitochondrion
39
. The depolarized
daughter mitochondrion is approximately six times less likely than the hyperpolarized daughter
mitochondrion to undergo a subsequent fusion event
39
. In this manner, fission generates a distinct
pool of low-ΔΨm mitochondria that are now segregated from the overall mitochondrial network of
the cell. Fission is thus a mechanism by which defective mitochondria can be compartmentalized
and subsequently removed from the mitochondrial network
38
. The process of mitochondrial fission
is mediated by the protein Drp1. Under normal conditions, Drp1 is localized to the cytoplasm.
During fission, it is recruited to the outer mitochondrial membrane, where it forms complexes that
join and then separate the outer and inner membranes
45
. In myoblast cultures, the presence of
excess fatty acids like palmitate in the medium leads to enhanced mitochondrial fragmentation
mediated by Drp1, and this is associated with a reduction in ΔΨm
46
. Inhibiting Drp1 in this model
prevents the fission-associated loss of ΔΨm
46
. This suggests that the recruitment of Drp1 to the
mitochondrial membranes could play a role in reducing ΔΨm.
Strikingly, Drp1 is markedly upregulated in invasive breast carcinoma and lymph node
metastases when compared to normal breast tissue
47
. Inhibiting Drp1 in breast cancer cells is
associated with a decrease in their in vitro migration and invasiveness
47
. Similar results are also
10
obtained with hepatocellular carcinoma cells, where mitochondrial fission is associated with cell
migration
48
. These findings suggest that fission may play an important regulatory role in the
development of metastases.
Mitophagy: In the final steps of ETC, molecular oxygen accepts the electrons shuttled
through respiratory complexes to form superoxide anions, which are then converted to hydrogen
peroxide and water by the subsequent actions of antioxidant enzymes. In instances of premature
electron leak from the ETC or dysregulation of certain mitochondrial enzymes, there is an
accumulation of ROS
45
. High levels of ROS are known to oxidize and damage mitochondrial DNA
as well as mitochondrial proteins including the respiratory proteins themselves
49
. From a quality
control perspective, it is important for cells to eliminate these mitochondria from the larger
mitochondrial network to prevent propagation of harmful mitochondrial DNA mutations or
damaged mitochondrial proteins. This process of selectively degrading defective mitochondria is
known as mitophagy (in other words, autophagy of mitochondria)
50
.
Depolarized mitochondria (mostly arising from fission with asymmetric distribution of
ΔΨm) are targeted for removal by mitophagy. Interestingly, depolarization of ΔΨm occurs long
before these mitochondria are eliminated by mitophagy, and inhibiting mitophagy does not recover
their ΔΨm
39
. Instead, it leads to an accumulation of low-ΔΨm mitochondria
39
. Mitophagy is thus
important to clear away the pool of low-ΔΨm mitochondria and maintain the overall cellular ΔΨm.
Studies have shown that the mechanism mediating mitophagy of low-ΔΨm specific mitochondria
involves the ΔΨm-dependent activation of PINK1-Parkin proteins
51
. Briefly, low-ΔΨm stabilizes
PINK1 which is present in the outer mitochondrial membrane
52
. Activated PINK1 then recruits
cytosolic Parkin
51,52
. The recruited Parkin then ubiquitinates and inactivates the mitofusin proteins,
thus isolating the low-ΔΨm mitochondria by preventing the mitochondria from undergoing fusion
11
and rejoining the overall mitochondrial network
51,53
. Parkin activation also leads to recruitment of
pro-mitophagy proteins such as p97
53
. Mitophagy is thus an important mitochondrial quality
control mechanism, and inhibition of mitophagy leading to accumulation of damaged
mitochondria has been implicated in promoting tumorigenesis
50,54
.
1.3.3: Mitochondrial and cellular motility: For cells to be motile and migrate in the
direction of certain cues, they need increased ATP production especially at the leading edges of
the migrating cell layer. Indeed, it has been reported that the cells at the leading edge of an
epithelial migrating sheet have higher ΔΨm
11
. A higher ΔΨm has also been associated with
increased migration in cancer cells
13
. Thus, the upregulation of ΔΨm may play an important role
in mediating cell migration. Additionally, even within individual cells, high-ΔΨm mitochondria
are transported selectively towards local regions of increased energy demand
36
. In neurons for
example, one study reported that mitochondria that travel from the cell body (soma) to the synapse
regions have high ΔΨm compared to the mitochondria that are transported from the synapse back
to the cell body
55
. Together, these results suggest that the ΔΨm is an important regulator of both
overall cell migration as well as individual mitochondrial motility within the cells.
1.4 Intrinsic regulation of the ΔΨm
ΔΨm regulates important mitochondrial as well as cellular processes including production
of ROS (which can serve as important cell signaling molecules when at physiological levels)
31
,
mitochondrial dynamics and quality control mechanisms, as well as mitochondrial and cell
motility. But what regulates the ΔΨm? In the following section we will examine different
12
intracellular mechanisms controlling the ΔΨm, from ion channels on the inner mitochondrial
membrane to regulation by the cytoskeleton and cell signaling pathways (Fig. 1-2).
Figure 1-2: Schematic showing intrinsic (top) and extrinsic (bottom) factors that could regulate
cancer cell ΔΨm.
13
1.4.1 Regulation of ΔΨm by inner membrane ion transporters: To ensure that the
electrochemical energy created by the proton gradient generated by the ETC is most effectively
used for ATP production, the inner mitochondrial membrane (IMM) must be mostly impermeable
to ions except through the ATP synthase. This would ensure that any ΔΨm dissipation is tightly
coupled with ATP production. Conversely, the flux of various ions across the IMM through their
respective ion channels (such as K
+
, Ca
2+
) can regulate ΔΨm and uncouple it from ATP
production
56
.
Mitochondrial calcium uniporter (MCU): These channels are the major source of Ca
2+
influx into the mitochondrial matrix
57
. ΔΨm provides the driving force for Ca
2+
flux through
MCUs; dissipating the ΔΨm with chemical uncouplers such as FCCP reduces Ca
2+
flux through
these channels
57-59
. Furthermore, ΔΨm itself is regulated by the MCU activity. Influx of Ca
2+
ions
leads to short-term ΔΨm depolarization
60,61
. Sustained high levels of Ca
2+
in the mitochondrial
matrix could induce the mitochondrial permeability transition pore (MPTP) and permanent loss of
ΔΨm
62
. Thus, MCUs can regulate ΔΨm by determining matrix Ca
2+
levels. Indeed, several studies
have shown that manipulating MCU levels can affect ΔΨm. In Drosophila S2 cells, knockdown
(k/d) of MCU rescued hydrogen peroxide induced ΔΨm-loss
63
. Interestingly, in liver cancer cells
it was found that k/d of MCUR1, a regulator of MCU that promotes mitochondrial Ca
2+
uptake, is
associated with a decrease in ΔΨm
64
. One possible explanation could be MCUR1 k/d leads to an
overall decrease in matrix Ca
2+
, which could reduce the activity of Ca
2+
-sensitive ETC enzymes
65
hence negatively affecting the H
+
pump that gives rise to the ΔΨm.
K
+
channels: There are several types of K
+
channels on the IMM: mitoKATP, mitoBKCa and
mitoKv1.3, which are regulated by ATP, Ca
2+
and voltage respectively
66
. These mitochondrial K
+
channels are thought to be involved in the regulation of ΔΨm, ROS generation as well as
14
mitochondrial matrix volume
67
. When there is ΔΨm hyperpolarization or an increase in ATP
concentration, the opening of these channels leads to the dissipation of the ΔΨm
68,69
and an
amelioration of the oxidative stress associated with enhanced ROS levels at high ΔΨm
67
. The ΔΨm
of cervical cancer cells overexpressing the pore-forming subunit of the mitoKATP channel is
significantly lower than that of the control cells, highlighting the ability of mitochondrial K
+
flux
in regulating ΔΨm
70
.
Uncoupling proteins (UCPs): The ETC Complexes I, III and IV and the major producers
of proton gradient; they pump protons into the mitochondrial intermembrane space. One of the
major consumers of this proton gradient is the ATP synthase (ETC Complex V) when it is
operating in its normal ATP-producing state. Over the past two decades, studies have found the
presence of yet another class of consumers of the proton gradient on the inner mitochondrial
membrane – the uncoupling proteins (UCPs)
71
. Originally found in brown adipose tissue (BAT),
UCP1 is activated by free fatty acids and transports protons into the mitochondrial matrix, thus
dissipating the ΔΨm without any resultant ATP production
72
. UCP1 is associated with energy
dissipation as heat, which is essential in non-shivering thermogenesis in BAT
72
. Later, other types
of UCPs were found to be expressed in several different cell types and associated with regulating
cell functions such as proton leak, ΔΨm, ROS levels, insulin secretion and resting metabolic rate
71
.
Notably, overexpression of UCP5 in neuroblastoma cells was found to decrease their ΔΨm
73
. In
aortic endothelial cells, manipulations that decreased or increased UCP levels led to an increase or
decrease in ΔΨm respectively
74
. These results provide compelling evidence that the ΔΨm can be
regulated by the level and activity of UCPs in cells.
15
1.4.2 Regulation of ΔΨm by outer membrane ion transporters: As mentioned earlier,
VDACs form channels that selectively allow passage of solutes through the outer mitochondrial
membrane based on their open/closed configuration. VDACs control mitochondrial function by
regulating the transport of important charged metabolites through the mitochondria (such as
succinate
2-
and ATP
4-
)
75
. Low ΔΨm lead to an ‘open’ configuration of VDAC which favors the
selective transport of anions over cations through these channels. Higher magnitudes of ΔΨm (>
40 mV) lead to a ‘closed’ configuration which favors the transport of cations
76
. Meanwhile, VDAC
closure could regulate ΔΨm by reducing the OMM permeability to adenine nucleotides (like
ANT)
77
. This would decrease substrate availability for the ΔΨm-consuming ATP synthase activity,
thus increasing the overall ΔΨm. Evidence for VDAC regulation of ΔΨm has been reported in
several studies. In neuroblastoma cells, dopamine stimulation leads to a transient decrease in ΔΨm,
which can be prevented by modulating the open/close configuration of the VDACs
78
. Moreover,
the ΔΨm-regulation by free cellular tubulin is also reported to be dependent on VDAC
phosphorylation; inhibiting VDAC phosphorylation by protein kinase A (PKA) abrogates the
inhibitory effect of enhanced free tubulin level on ΔΨm
79
. Further investigation on VDAC isoforms
shows that, while inhibition of different VDACs (using siRNA) all leads to a decrease in ΔΨm,
VDAC3 k/d causes the greatest ΔΨm reduction
76,80
. Additionally, VDAC activity can also be
regulated by binding to hexokinase, a cytoplasmic enzyme involved in glycolysis
81
. K/d of
hexokinase using shRNA decreases the ΔΨm of cancer cells, an effect that can be reversed by the
presence of VDAC inhibitors
82
.
1.4.3 ΔΨm regulation by cytoskeletal elements: The cytoskeleton regulates the spatial
organization of subcellular organelles and couples them biochemically and physically to the
16
extracellular cues
83
. It is mainly comprised of three types of polymers: microtubules,
microfilaments (or actin filaments), and intermediate filaments
83
. Studies have shown that the
cytoskeletal filaments can control mitochondrial motility, morphology and subcellular location
84
.
In this section, we will examine the role of these cytoskeletal filaments in regulating the ΔΨm.
Microtubules: Microtubules are one of the main cytoskeletal elements associated with the
transport of mitochondria within the cells
84
. They are the stiffest cytoskeletal element
83
and are
made of tubulin heterodimers. Microtubule polymerization can be blocked by drugs that act as
microtubule destabilizers such as nocodazole, rotenone and colchicine
85
, which increase levels of
free tubulin in cells
85
. On the other hand, microtubule polymerization can be stabilized or enhanced
by treatment with drugs such as paclitaxel, which decreases free tubulin levels as shown in
hepatocellular carcinoma cells
85
. Interestingly, microtubule depolymerization (increased free
tubulin) decreases ΔΨm while its stabilization (decreased free tubulin) increases ΔΨm in cancer
cells
85
. This contrasts with non-cancer cells where microtubule stabilization does not affect ΔΨm
85
.
Other studies show that both depolymerizing and stabilizing microtubule decrease ΔΨm in
myocytes
86
. In neurons, microtubule stabilization with paclitaxel leads to a significant decrease in
the ΔΨm
87
. In human dermal fibroblasts, depolymerizing microtubule with nocodazole does not
change ΔΨm
88
. Together, these studies show that the microtubule control of ΔΨm is highly cell-
type specific. A potential mechanism by which microtubule stability affects ΔΨm involves the
binding of free tubulin to and closing of VDACs in the outer mitochondrial membrane
89
, which
reduces ADP transport into the mitochondria and overall mitochondrial activity
85,89
.
Actin filaments (or microfilaments): In contrast to microtubules which mediate long-
distance transport of mitochondria within the cells, actin filaments have been associated with short-
range mitochondrial transport as well as mediating the anchorage of the mitochondria in regions
17
of high energy demand
90
. Monomeric actin (G-actin) can polymerize and form diverse structures
of filamentous actin (F-actin)
83
. The activity of several types of actin regulatory proteins can
dictate the overall structure of the cell’s actin network
83
. For example, actin assembly in the
presence of formins leads to the formation of liner actin filaments, whereas the activity of the
Arp2/3 complex leads to branched actin networks
91
. In T-cells and neurons, gelsolin (an actin
severing protein that caps actin filaments and prevents their elongation) is reported to prevent the
loss of ΔΨm that is associated with the onset of apoptosis
92,93
. Additionally, a small fraction of the
overexpressed gelsolin in T-cells was found to colocalize with the mitochondria, suggesting that
it may regulate ΔΨm by direct binding
92,94
. Interestingly, a more detailed study revealed that the
mitochondrial fraction of the overexpressed gelsolin specifically bound to the VDACs, suggesting
a regulatory mechanism of ΔΨm by gelsolin through VDACs
95
. Coronin, another actin binding
protein that opposes the activity of Arp2/3 and thus inhibits F-actin formation is essential for
maintaining ΔΨm in T-cells. Loss of Coronin-1 leads to a decrease in ΔΨm, which can be partially
recovered by treatment with latrunculin A (an actin depolymerizing drug)
96
. On the other hand,
jasplakinolide treatment, which increases F-actin, leads to a decrease in ΔΨm
96,97
. These results
suggest that the ratio of G-actin to F-actin in cells could potentially regulate their ΔΨm. A recent
study in cancer cells reported that at any given time, about 20% of the mitochondria are associated
with actin filaments (F-actin), and this association while regulating mitochondrial dynamics, had
no effect on their ΔΨm
98
. Regulation of ΔΨm by the actin cytoskeleton could therefore be highly
context dependent and cell-type specific.
The relationship between actin dynamics and mitochondrial activity has also been explored
in other model organisms such as yeast and plant cells. In yeast cells, mutations in actin regulatory
proteins that promote the turnover of F-actin (Sla1p and End3p) lead to a loss of ΔΨm
99
. In plant
18
root hair cells, both actin polymerizing and depolymerizing treatments (with jasplakinolide and
latrunculin B respectively) reduce the ΔΨm
100
. In agreement with these results, latrunculin
treatment in mung bean plant cells also reduces the ΔΨm
101
. Taken together, these studies provide
compelling evidence that the dynamics of cellular actin cytoskeleton could regulate ΔΨm.
Intermediate Filaments (IFs): IFs have the least stiffness out of the three types of
cytoskeletal elements and are thus better suited for receiving mechanical stimuli of tensile rather
than compressive nature
83
. Mutations in IF genes are associated with altered mitochondrial
distribution and morphology in several pathological conditions including Charcot-Marie-Tooth
disease and epidermolysis bullosa simplex
102
. In mouse fibroblasts, a portion of vimentin IFs
colocalizes with mitochondria
103
. Further, disrupting the vimentin IFs leads to a redistribution of
the mitochondrial network as well as increased mitochondrial motility, suggesting that vimentin
IFs are involved in regulating the spatial organization and motility of mitochondria within these
cells
103
. Interestingly, motile mitochondria were found to have a lower ΔΨm than their stationary
counterparts
104
; restoring vimentin expression in the vimentin-null fibroblasts decreased
mitochondrial motility
103
while increasing their ΔΨm
104,105
. Conversely, silencing vimentin using
shRNA in wild-type cells (which express normal levels of vimentin) leads to a decrease in ΔΨm
105
.
A more detailed study revealed that the vimentin IF-mitochondria association could be disrupted
by the phosphorylation of vimentin at its Ser-55 site by Rac-1, and activated Rac-1 increases
mitochondrial motility and reduces the ΔΨm
102
. Together, these results strongly suggest that the
association of mitochondria with vimentin IFs helps anchor the mitochondria at specific
intracellular locations (thus reducing their motility) and increases their ΔΨm.
19
1.4.4 Regulation of ΔΨm by signaling pathways: The activities of several mitochondrial
proteins, including the ETC complexes, can also regulate the ΔΨm. These activities can be
expression of tissue-type specific isoforms of ETC subunits, their allosteric inhibition from
interactions with small molecules (such as ADP), and their altered phosphorylation state by cell
signaling molecules
106
. For example, both ETC Complexes I and IV, which contribute to ΔΨm
through proton pumping, are regulated via PKA-mediated phosphorylation induced by c-AMP
107
.
Akt was found to localize to the mitochondria following PI3K signaling, to phosphorylate ATP
synthase (ETC Complex V)
108
. Whether the phosphorylation of ETC Complexes serves to activate
or inhibit them seems to be dependent on the subunit type and site of phosphorylation. Notably,
several other signaling molecules, including tyrosine kinases such as Abl, Src, EGFR and ErbB2
and serine threonine kinases like JNK, GSK3β, were also found to localize to the mitochondria
and modulate ETC Complex activity
109
. As such, various upstream signaling pathways can
potentially regulate the ΔΨm.
1.5 How does the TME regulate ΔΨm of cancer cells?
The immediate environment surrounding tumors in vivo, commonly referred to as the
tumor microenvironment (TME), is a rich source of biophysical and biochemical cues that directly
influence cancer cell behaviors. Several different cell types apart from the primary tumor cells are
found in the TME: immune cells, fibroblasts, adipocytes, mesenchymal stem cells, endothelial
cells and pericytes
110
. Biochemical cues derived from cell-type specific signaling are important in
directing tumor progression. For example, the interaction of cancer cells and dendritic cells in the
TME is thought to regulate Treg expansion within the tumor and thereby mediate tumor escape
from immune attacks
111
. In addition, tumor progression in the TME has been associated with
20
biophysical cues such as low oxygen concentrations (hypoxia)
112,113
, remodeling of the
extracellular matrix (ECM) resulting in a change in both its composition and stiffness
114
, as well
as compressive and shear stresses
115,116
. These TME-derived cues are potential regulators of cancer
cell mitochondrial activity and their roles in influencing the ΔΨm are explored below (Fig. 2).
1.5.1: Hypoxia: Exposure to hypoxia induces several changes in mitochondrial metabolism
and the ΔΨm response to hypoxia appears to be highly cell-type specific and metabolic state-
dependent. In myoblasts
117
, infected macrophages
118
and kidney proximal tubule cells
119
, hypoxia
has been reported to depolarize the ΔΨm
120
. However, in blastocysts
121
and hypoxia-sensitive
glioblastoma cells
122
, the ΔΨm increases in response to hypoxic stimuli. Furthermore, in hypoxia-
resistant glioblastoma cells
122
, uninfected macrophages
118
and renal cell carcinoma cells
123
, the
ΔΨm remains unchanged under hypoxia. One potential explanation for these differences is in the
distinct ways individual cell types adapt the efficiency and rate of mitochondrial respiration under
hypoxia. In hypoxia-resistant glioblastoma cells, the ΔΨm remains constant under hypoxia, likely
due to a compensatory decrease in overall oxygen consumption or stabilization of the ETC, leading
to constant proton flux into the mitochondrial intermembrane space despite the reduced electron
flux, which is not seen in the hypoxia-sensitive glioblastoma cells
122
. Additionally, hypoxia has
been associated with several other mechanisms that could determine the ΔΨm response. These
include HIF-1 induced inhibition of pyruvate dehydrogenase, which decreases the overall
mitochondrial respiration by reducing substrate availability, expression of hypoxia-specific
isoforms of ETC complexes (cytochrome c oxidase), inhibition of ETC complex I, and the
inhibition of ATP-synthase.
21
1.5.2 Substrate Stiffness and Composition: Stromal secretion of lysyl oxidase (LOX)
enzymes in the TME has been associated with increased ECM stiffness, promoting cancer cell
migration and invasion
114
. Substrate stiffness has been shown to regulate the ΔΨm of several
different cell types in vitro. Pulmonary arterial endothelial and smooth muscle cells
124
, and
cardiomyocytes
125
demonstrate enhanced ΔΨm when cultured on softer substrates. Another study
on vascular smooth muscle cells found that their ΔΨm was the highest when cultured on a substrate
with stiffness close to the in vivo stiffness of the tissue of origin
126
, while both softer and stiffer
substrates significantly reduced the ΔΨm of these cells. In addition to ECM stiffness, the overall
ECM composition can also be modified in the TME which may in turn alter ΔΨm. In several tumor
types, there is an increased deposition of fibrillar collagen, fibronectin and hyaluronan, leading to
a desmoplastic phenotype that is linked to poor prognosis
114
. It was shown that pancreatic cancer
cells cultured on substrates coated with fibronectin and laminin have a much higher ΔΨm than
those on collagen coated or non-adherent substrates
127
. Furthermore, altering the composition of
anabolic vs. catabolic ECM proteins in nucleus pulposus cells was also found to change their
ΔΨm
128
. In vascular smooth muscle cells, silencing an ECM protein (cartilage oligomeric matrix
protein) led to a decrease in their ΔΨm
129
. Together, these studies suggest a possible role of ECM
stiffness and composition in regulating cancer cell ΔΨm.
1.5.3 Compressive and tensile stresses: As tumors grow, they push against the surrounding
stromal tissue and accumulate solid stresses within the tumor. It was shown that there are
compressive forces at the tumor core and tensile stresses at the tumor periphery
130
. While the
compressive stress has been reported to enhance the migration of breast cancer
131
and glioma
132
cells, its effect on ΔΨm is still unclear and remains to be investigated. On the other hand, tensile
22
stresses have found to regulate ΔΨm in various cell types in vitro. In human neuroblastoma cells,
mild to moderate stretch was shown to depolarize the ΔΨm
133
. Interestingly, sustained mechanical
stretch reduces the ΔΨm of cardiomyocytes but has no effect on that of cardiac fibroblasts
134
.
Interestingly, bovine aortic smooth muscle cells have increased ΔΨm under both monotonous and
variable stretching
135
. These data demonstrate a potential role of static and dynamic tensile stresses
in TME in altering the ΔΨm in cancer cells.
1.6 Techniques to investigate ΔΨm
1.6.1 In vivo techniques: Determining the ΔΨm of cells in vivo is not very straightforward
and remains challenging through mitochondrial dyes and direct imaging. Some recent studies have
reported the use of novel approaches to measure the ΔΨm of live cells in vivo. Lipophilic cations
such as triphenylphosphonium (TPP) can readily accumulate in the mitochondria (owing to the
ΔΨm), which can be conjugated with compounds for mitochondrial targeting
136
. A novel TPP-
based approach was reported by Logan et al., where two TPP-based probes with complementary
click chemistry moieties are injected in vivo (in mouse models) that accumulate in the
mitochondria in a ΔΨm (and plasma membrane potential) dependent manner, where they form a
stable “MitoClick” compound. The amount of “MitoClick” accumulated in different cell types or
tissues can be quantified by using liquid chromatography-tandem mass spectrometry (LC-
MS/MS), and this approach was shown to sensitively report small changes in ΔΨm in vivo
137
. The
second TPP-based approach to measure ΔΨm in vivo uses a radiolabeled TPP tracer that
accumulates in tissues in a ΔΨm-dependent manner and can then be imaged using positron
emission tomography (PET). Using this approach in a mouse model of lung tumor, Momcilovic et
al. report that the lung tumors could be segregated into distinct subtypes (adenocarcinoma vs.
23
squamous cell carcinoma) based on their estimated in vivo ΔΨm
25
. These TPP-based approaches
are promising new techniques for in vivo estimation of ΔΨm.
1.6.2 In vitro techniques: The most widely used techniques to monitor the ΔΨm in vitro
are based on lipophilic cationic fluorescent dyes that accumulate in the mitochondria in a ΔΨm-
dependent manner
138
. The ΔΨm can then be assessed by quantifying the fluorescence of these
mitochondria-accumulated dyes using either fluorescence microscopy, flow cytometry or
microplate reader-based approaches. Some of the commonly used ΔΨm fluorescent probes include
tetramethylrhodamine methyl ester (TMRM), 5,5′,6,6’-tetrachloro-1,1’,3,3′-
tetraethylbenzimidazolylcarbocyanine iodide (JC-1), 1,1′,3,3,3′,3′-hexamethylindodicarbo -
cyanine iodide (DiIC1(5)), and rhodamine 123 (Rho123)
138,139
. Out of these dyes, TMRM is
known to have the least inhibitory effect on mitochondrial activity and hence is the most preferred
dye to use for in vitro ΔΨm measurements
138
.
1.7 Objectives and aims
Cellular ΔΨm is an important indicator of mitochondrial activity and regulates a wide range
of cell functions including apoptosis and migration. Most cells derived from epithelial carcinomas
are known to have an abnormally high ΔΨm
10,11
, which has been associated with altered behaviors
such as susceptibility to apoptosis
12
and migratory potential
13
. However, the source of this
differential regulation of ΔΨm in cancer cells is unclear. We will therefore investigate how cues
from the tumor microenvironment (TME) can alter cancer cell-ΔΨm.
24
Central Hypothesis: Specific cues from the TME can alter the ΔΨm of cancer cells, which
plays an important role in altering their invasive and migratory properties. Understanding how the
TME regulates ΔΨm will lead to novel targets for developing cancer therapeutics.
Objective: To use engineered platforms to model and investigate how the TME regulates
the ΔΨm of cancer cells.
Aim 1: Utilize an engineered micropatterning platform to model the differential regulation and
heterogeneity of cancer cell ΔΨm in the TME.
We aimed to recapitulate the heterogeneity of cancer cell ΔΨm in engineered in vitro
models to provide novel insights into ΔΨm-altering TME cues. Specifically, we micropatterned
tumor islands surrounded by stromal cells that show a spatial ΔΨm distribution and used laser
capture microdissection to extract cancer cells from specific locations within the micropatterns and
assessed differences in their gene expression using RNA sequencing. We also investigated the
correlation of spatial ΔΨm levels with transcription factor activation and targets that perturb the
ΔΨm distribution. Further, we also look at whether cancer cells sorted based on their ΔΨm and
injected into mouse models, lead to differences in metastatic lung burden. The results from this
study are presented in Chapter 2.
Aim 2: Investigate how differential E-cadherin adhesion cues from the TME regulate the ΔΨm of
cancer cells.
In this study, we used an in vitro micropatterning platform to recapitulate biophysical
confinement cues in the TME and investigated the mechanisms by which these regulate cancer
cell ΔΨm. We found that micropatterning resulted in a spatial distribution of ∆Ψm, which
25
correlated with the level of E-cadherin mediated intercellular adhesion. There was a stark contrast
in the spatial distribution of ΔΨm in the micropattern of E-cadherin-negative breast cancer cells
(MDA-MB-231) compared to that of the high E-cadherin expressing (MCF-7) cancer cells.
Disruption and knockout of E-cadherin adhesions rescued the low ΔΨm found at the center of
MCF-7 micropatterns with high E-cadherin expression, while E-cadherin overexpression in MDA-
MB-231 and MCF-7 cells lowered their ΔΨm at the micropattern center. These results show that
E-cadherin plays an important role in regulating the ΔΨm of cancer cells in the context of
biophysical cues in TME and are presented in Chapter 3.
Aim 3: Investigate how downstream effector transcription factor YAP regulates ΔΨm.
In this study we aimed to investigate the downstream effectors of E-cadherin mediated
regulation of ΔΨm. We hypothesized that since cytoplasmic YAP correlated with low-ΔΨm in
tumor micropatterns, it could play a role in negatively regulating cancer cell ΔΨm. We showed
how YAP overexpression decreases cancer cell ΔΨm and mitochondrial respiration while
increasing glycolysis. We also found that YAP knockdown led to an increase in mitochondrial
transcription, indicating a mechanism by which YAP could negatively regulate mitochondrial
activity. Furthermore, we also found the presence of YAP in mitochondrial protein isolates. These
results suggest a mitochondrial localization of YAP where it behaves as a transcription repressor
and thus functions as an important regulator of cancer cell ΔΨm. These results are presented in
Chapter 4.
26
1.8 Novelty and Impact
It has been known for decades that cancer cells have abnormally high ΔΨm. However, it
remains unclear whether this phenotype is a result of cell-intrinsic mutations specific to and
characteristic of cancer cells, or due to the activation of aberrant intracellular signaling pathways
or from extrinsic cues in the TME. Given the crucial role of the TME in dictating cancer cell
behavior, we propose that specific cues from the TME can alter the ΔΨm of cancer cells through
intracellular physicochemical signaling. A better understanding of the source of altered cancer cell
ΔΨm can lead to the discovery of new targets for attacking cancer cells, by reducing their metabolic
plasticity and potentially their metastatic ability. Since the abnormally high ΔΨm has been
observed in a wide variety of epithelial carcinoma cells, a detailed understanding of the underlying
mechanisms giving rise to this phenotype will be beneficial for developing new therapeutic targets
for a wide range of carcinomas.
27
Chapter 2: Spatial Regulation of Mitochondrial Heterogeneity by
Stromal Confinement in Micropatterned Tumor Models
2.1 Rationale
Despite the latest advances in cancer therapeutics, the five-year survival rate for patients
diagnosed with metastatic breast cancer is at a staggering 27% as opposed to 99% for those with
localized disease
140
. A better understanding of metastatic progression is therefore crucial to
improve prevention and treatment of advanced breast cancer.
Mitochondria have recently emerged as a potential regulator of cancer progression and
metastasis. It has been shown that cancer cells depend on mitochondrial respiration for their in
vivo tumor-forming ability
141-143
, and that mitochondrial metabolites play a role in driving
oncogenesis
144
and epithelial-mesenchymal transition (EMT)
145
, a phenotypic switch that precedes
metastasis
146
. Importantly, there is significant heterogeneity in mitochondrial phenotypes across
cancer disease stages, and even in the same patient. Increased mitochondrial redox activities in
tumors have been correlated with tumor aggressiveness and metastatic potential
147
. Higher
mitochondrial membrane potential (ΔΨm) is associated with cancer cell survival and
invasiveness
13,148-150
. In breast cancer, circulating tumor cells (CTCs), the presumptive precursor
of metastases, exhibit enhanced mitochondrial biogenesis and respiration compared to cancer cells
from primary tumors in the same host
151
. However, questions regarding where and how the
heterogeneity of mitochondrial activities arises, and its impact on metastatic development, remain
unanswered.
The tumor microenvironment (TME) plays an important role in cancer progression and
metastasis
152
. The TME of progressing breast tumors is often characterized by distinct architectural
and cytological features, including an evolving tissue interface of direct tumor-stromal
28
interactions
153-156
and a stiffening tumor mass
130
. Recently, it was reported that some stromal cells
can regulate metabolic and/or mitochondrial functions in cancer cells through paracrine growth
factor signaling and metabolite exchange
157,158
, or through transfer of mitochondrial DNA into
cancer cells
142,143
. On the other hand, biomechanical properties of the TME have also been found
to influence cancer cell invasiveness and metastatic potential
130,159
. At the tissue level, mechanical
stresses in solid tumors are spatially dependent on tumor architecture and growth
160
. At the cellular
level, mechanical cues are involved in regulating cancer cell proliferation
115
, invasiveness
161
, and
extracellular matrix (ECM) remodeling
131,162
. Although not yet reported in cancer cells, it has been
shown that mechanical stimuli can affect mitochondrial activity in cardiomyocytes and endothelial
cells
126
, and that cytoskeletal remodeling leads to changes in mitochondrial dynamics
98
. However,
it remains unclear whether stromal cells and their associated mechanical cues within the tumor
architecture can drive heterogeneous mitochondrial activities.
We make use of a previously established a micro-engineered tumor model, i.e., a
micropatterned tumor-stromal assay (μTSA), to demonstrate that tumor-stromal interactions
within the architectural context of a tumor play an important role in inducing phenotypic
heterogeneity in cancer and stromal cells in vitro, which was confirmed in vivo
155
. In this study,
we use the µTSA to investigate the role of stromal cells in regulating tumor mitochondrial
heterogeneity in vitro. We reveal that ΔΨm, mitochondrial mass, metabolism, and metastatic
potential are spatially distributed within the µTSA. We show that ΔΨm levels correlate with
YAP/TAZ nuclear translocation status, are regulated by stromal confinement, and are dependent
on actin cytoskeleton. We further demonstrate a positive correlation between ΔΨm levels and
metastatic potential in cancer cells in vivo. We demonstrated in the µTSA a new regulatory role
that the TME plays in mitochondrial heterogeneity and metastatic potential.
29
2.2 Materials and Methods
2.2.1 Cell Culture
MCF-7 and MDA-MB-231 cells (ATCC) were cultured in Dulbecco’s Modified Eagle
Medium (DMEM; Life Technologies) supplemented with 10% Fetal Bovine Serum (FBS; EMD
Millipore), and 100 U/ml penicillin and 100 µg/ml streptomycin (Life Technologies). Primary
human bone marrow stromal cells (BMSCs) derived from whole human bone marrow aspirates
(Lonza)
163
were expanded using the MesenCult Proliferation Kit (Stem Cell Technologies). µTSA
co-cultures were carried out in the FBS and penicillin/streptomycin supplemented DMEM.
2.2.2 Antibodies
The following primary antibodies were used in this study: TOM20 (Santa Cruz
Biotechnology, sc-17764, 1:50), YAP/TAZ (Cell Signaling, rabbit mAb 8418, 1:200), pan-keratin
(C11 mouse mAb 4545, 1:500), and vimentin (Cell Signaling, rabbit mAb 5741, 1:200). Alexa
Fluor-conjugated secondary antibodies (Life Technologies) were used at a dilution of 1:500.
2.2.3 Micropatterned Tumor-Stromal Assay (μTSA)
Collagen-I was extracted from rat-tails (acetic acid extraction)
164
and quantified using the
modified Lowry method
165
. Glass coverslips (12 mm) were cleaned with 7X cleaning solution (MP
Biomedicals), incubated with plasma, silanized with 3-aminopropyltriethoxysilane (APTES), and
coated with 0.1mg/ml extracted collagen. A laser engraver (Epilog) was used to print stencils on
250 µm thick silicone sheets (PDMS), which were then successively cleaned in 70% isopropyl
30
alcohol and Milli-Q water, and air-dried. Cleaned stencils were then aligned on top of collagen-
coated coverslips, and this assembly was treated with 0.2% pluronic F-127 (Sigma) for 15 minutes,
followed by PBS and DMEM rinses. The stencil-overlaid coverslips were placed in 24-well plates.
MCF-7 cells were seeded at a density of 300,000 cells per micropattern in the 24-well plates and
allowed to adhere onto the circular islands within the micropatterns for 5 hours. Next, the
micropatterns were rinsed in DMEM to remove excessive cancer cells. After overnight incubation,
stencils were removed from the coverslips and micropatterns were rinsed thoroughly in DMEM.
BMSCs were seeded (at densities ranging from 25,000 to 100,000 cells per micropattern) and
rinsed off from the cancer island by thorough DMEM rinsing within 30 minutes of seeding to
allow for adhesion in the surrounding areas without contaminating the cancer cell region (day 0).
The micropatterned cells were cultured for 4 days before the subsequent analysis.
2.2.4 Imaging of the Mitochondrial Membrane Potential (ΔΨm) in the μTSA
Day 4 micropatterns were stained at 37
o
C for 30 minutes with the mitochondrial membrane
dye, tetramethylrhodamine methyl ester, (30 nM in complete medium; TMRM, Life
Technologies). Following incubation, micropatterns were rinsed and immediately imaged in PBS
using a Nikon Eclipse Ti inverted fluorescence microscope.
2.2.5 Immunofluorescence
Day 4 micropatterns were fixed with 4% paraformaldehyde (Electron Microscopy
Sciences) at room temperature for 15 minutes, rinsed with PBS, and permeabilized with 0.1%
Triton X-100 (Fisher Scientific) for 10 minutes. Micropatterns were then blocked with 4% bovine
31
serum albumin (GE Healthcare), either for one hour at room temperature or overnight at 4
o
C,
followed by incubation with the primary antibody for 2 hours at room temperature, three PBS
rinses and incubation with the respective secondary antibodies at room temperature for 1 hour.
Micropatterns were then rinsed three times with PBS and once with milli-Q water, mounted onto
glass slides using FluoroGel II (Electron Microscopy Sciences), and imaged with the Nikon
Eclipse Ti inverted microscope and/or the Nikon confocal microscope.
2.2.6 Metabolic Imaging (NAD(P)H and FAD fluorescence)
A Zeiss LSM-780 inverted confocal microscope coupled to a Ti-Sapphire laser system and
an A320 FastFLIM FLIMbox
166
was used to determine the fluorescence intensities of NADH and
FAD in day 4 MCF-7-BMSC μTSA. Image processing and quantification were performed using a
customized code in Python2.7.
2.2.7 Laser Capture Microdissection (LCM) and RNA Sequencing
Day 4 micropatterns were fixed in 100% ethanol, serially rehydrated in 75% ethanol and
water, and stained with the Histogene Staining Solution (Applied Biosystems). The samples were
then serially rinsed and dehydrated in water, 75%, 95%, and 100% ethanol, submerged in xylene,
and air-dried immediately before being mounted onto glass slides with cell side up for LCM. The
ArcturusXT Laser Capture Microdissection System was used to extract MCF-7 cells at the centers
and edges of the MCF-7-BMSC µTSA. RNA from the micro-dissected cells was extracted using
the PicoPure RNA Isolation Kit (Arcturus) following the manufacturer’s instructions. RNA
concentrations were determined by absorbance on a NanoDrop One spectrophotometer and Qubit
32
(ThermoFisher). RNA integrity was determined using the Agilent 2100 Bioanalyzer and RNA
6000 Nano Chip kit. Only samples with an RNA Integrity Number (RIN) score greater than 7 were
used. Following mRNA purification from total RNA using poly-A selection, a cDNA library was
prepared using Illumina TruSeq Stranded mRNA Library Prep kit. The cDNA library quality was
checked using the Agilent 2100 Bioanalyzer and D1000 DNA Chip kit. Samples were run on the
Illumina NextSeq 500 platform using the High Output Sequencing Kit v2 (150 cycles, 2x75bp read
length, 20 million reads per sample). Library preparation and sequencing were performed by the
Single Cell, Sequencing, and CyTOF (SC2) Core (Children’s Hospital of Los Angeles, Los
Angeles, CA).
2.2.8 Gene Set Enrichment Analysis (GSEA)
Before analysis, raw FASTQ files from RNA-sequencing were checked for read quality
using FASTQC. Next, reads were mapped to the most recent Homo sapiens reference genome
(GRCh38) using the HISAT2 splice aligner
167
. After alignment, mapped reads were counted with
R using the Rsubread Bioconductor package
168
. Read counts were arranged into a count matrix
and differentially expressed gene (DEG) analysis was performed using DESeq2 in R
169
. A pairwise
comparison using the Wald Chi-Squared test was employed to contrast gene expression within our
two groups (μTSA interface versus center) and genes were subsequently ranked based on
comparison significance (-log10(nominal p-value)*LFC direction; where LFC is the log fold
change). GSEA was then conducted under default settings to identify the coordinated enrichment
of functionally linked genes
170,171
. Briefly, the GSEA algorithm walks through the pre-ranked
dataset and calculates a running enrichment score (ES) by increasing the ES when a dataset gene
is found to be a member of the gene set of interest and decreasing the ES if it is not. A gene set is
33
identified as enriched when the final ES is high and weighted towards a single phenotype.
Statistical significance is determined by comparing the ES to the ESnull obtained by permutating
the ranked dataset 1000 times and recalculating the ES. Nominal p-values are then adjusted for
multiple testing correction through the false discovery rate (FDR) Benjamini-Hochberg method
172
.
Gene sets with an FDR < 0.25 were considered as significantly enriched. Functionally linked gene
sets were obtained from the Hallmark
173
, KEGG, and Gene Ontology
174
collections compiled in
Molecular Signatures Database (MSigDB)
170
.
2.2.9 Mechanical Constraints
A microcontact printing approach was used to create a PDMS barrier for the complete
mechanical constraint in µTSA. To create a mold for the stamps, circular outlines (2 mm in
diameter) were cut on a 140 µm thick protective film tape (Patco) with a desktop craft cutter
(Silhouette Cameo). The circular tape cutouts were then transferred onto the bottom of 150 mm
diameter petri dishes, and covered with a 1 mm-thick layer of PDMS (curing agent:base = 1:10,
Sylgard 184, Dow Corning), degassed, and cured at 65
o
C. The cured PDMS layer was then cut
with a biopsy punch (10 mm in diameter) to obtain PDMS stamps with the 2 mm concave circular
region at their center replica-molded from the tape cutouts. To create a PDMS barrier, a thin layer
of freshly mixed and degassed PDMS was spin-coated on a round 18 mm-diameter glass coverslip.
A stamp was placed on top of the thin PDMS layer, inked with PDMS, and then transferred onto
a collagen-coated coverslip. The stamp was then removed and discarded. The transferred PDMS
layer was cured overnight at room temperature. Before use, the coverslips were incubated in 0.2%
pluronic F-127 (Sigma) for 15 minutes followed by rinses with PBS and DMEM. MCF-7 cells
were then seeded to create completely confined tumor micropattern.
34
2.2.10 In vivo Metastasis Model
All animal studies were carried out in accordance with the recommendations in the Guide
for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was
approved by the Institutional Animal Care and Use Committee of the University of Southern
California. MCF-7 and MDA-MB-231 breast cancer cells (ATCC), stably transduced with a
GFP/luciferase construct
175,176
, were stained with TMRM or DiIC1(5) (MitoProbe DiIC1(5) Assay
Kit, Invitrogen) as a marker of mitochondrial membrane potential and sorted on the BD FACSAria
I cell sorter (Flow Cytometry Core Facility, USC). Sorted cells were inoculated into NSG
(NOD.Cg-Prkdc
scid
Il2rg
tm1Wjl
/SzJ) mice through tail vein injections (300,000 cells per mouse for
MCF-7 and 100,000 cells per mouse for MDA-MB-231). For the MCF-7 study, mice were
implanted with 17β-estradiol 90-day release pellets (0.36mg/pellet), using a sterile trochar 24
hours prior to the injections. After five weeks (for MCF-7 cells) or four weeks (for MDA-MB-231
cells), the animals were euthanized and their lungs were harvested, ground, and assayed for
metastatic tumor content using a Luciferase Assay System (Promega)
155
2.2.11 Statistical Analysis
Other than the RNA sequencing data (see above), all data are presented as mean ± standard
deviation (SD), as stated in the figure legends. Statistical significance was assessed using the
Welch’s t-test (parametric) and Mann-Whitney (non-parametric) for pair-wise comparison, and
ordinary 1-way ANOVA for comparison between multiple (≥3) conditions; p<0.05 was considered
as significant.
35
2.3 Results
2.3.1 Mitochondrial and metastatic pathways are upregulated at the tumor-stromal interface
in the µTSA
To recapitulate cell-cell interactions within the tumor architectural constraints of the
TME(Shen et al. 2014), MCF-7 breast cancer cells were cultured in the µTSA with surrounding
bone marrow stromal cells (BMSC)
177
(Fig. 2-1A), resulting in adjacent spatial compartments of
cancer and stromal cells with a defined tumor-stromal interface (Fig. 2-1B). To examine whether
cancer cell phenotypes and signaling were differentially regulated at the transcriptome level, we
used laser-capture microdissection (LCM) to obtain cancer cells from the interfacial and central
regions of the µTSA island, and performed RNA sequencing (RNA-seq) on these cells. The RNA-
seq data were analyzed by gene set enrichment analysis (GSEA), where statistical values were
generated through the Wald Test, and gene sets were ranked by statistical significance and fold
change direction (see Methods). Among the pathways that differ between the interface and center,
we identified three categories gene sets among the 8 that were of interest in this study (Fig. 2-1C),
i.e., 1) mitochondrial proton/electron transport activities; 2) cellular metabolism; and 3) cancer
invasion and metastasis
178-185
. Specifically, genes related to mitochondrial activities were highly
dependent on the cell location in the µTSA, with cells at the interface exhibiting upregulation of
the electron and proton transport genes. At the metabolic level, interfacial cancer cells had
upregulation of genes involved in oxidative phosphorylation (OXPHOS), contrary to cells at the
center of the micropattern that showed upregulation of glycolysis genes. Consistent with the
OXPHOS upregulation, genes from the PI3K-Akt-mTOR pathway
186
were found to be upregulated
at the interface. Notably, gene sets involved in actin-cytoskeleton and TME-mediated invasiveness
were also upregulated at the interface, suggesting a role of the interface in cancer cell mechanics
36
and migration. Importantly, a gene set that has been found to predict metastasis in lymph node-
negative ER+ breast cancer
184
was also upregulated at the interface (Fig. 2-1C), matching the ER+
subtype of MCF-7. The results here demonstrate spatially regulated heterogeneity of gene sets for
mitochondrial activities, metabolism, and metastatic potential in cancer cells in the µTSA.
Figure 2-1. Spatial regulation of mitochondrial, metabolic, and metastatic pathways in
micropatterned tumor-stromal assays (µTSA) assessed by RNA sequencing. (A) Schematics
depicting the steps to create a MCF-7/BMSC µTSA; (B) Representative micropattern of an MCF-
7 island (green: Pan-Cytokeratin) surrounded by BMSCs (red: Vimentin). Blue: nuclei. Scale bar:
100 μm. Cancer cells from the center and edge of the µTSA (opaque annular ring with dotted
outlines) were isolated by laser capture microdissection
155
, and their RNA was extracted and
sequenced; (C) Gene sets found to be enriched at the interface vs. center by RNA sequencing and
37
gene set enrichment analysis (glycolysis has negative NES indicating enrichment in the center).
(See Methods for more details. FDR: False Discovery Rate; FDR < 0.25 was considered
significant.) MCF-7 cells from > 6 micropatterns/experiment were microdissected and combined
before RNA isolation and sequencing; N = 2 independent experiments.
2.3.2 Mitochondrial membrane potential (ΔΨm) is spatially regulated in the µTSA
To confirm that the mitochondrial proton/electron transport pathway is upregulated in the
µTSA, we stained cells with tetramethylrhodamine methyl ester (TMRM), a reversible
mitochondrial membrane potential (ΔΨm) dye, which only accumulates in active mitochondria
with intact membrane potential
187
. As shown in Fig. 2-2A, the ΔΨm of MCF-7 cells in the µTSA
had a distinct spatial profile, with lower ΔΨm in cells at the center of the tumor micropattern, and
higher in those closer to the interface. To validate that the TMRM signal is reflective of ΔΨm, we
treated the micropatterns with the ΔΨm uncoupler, carbonyl cyanide-4-
(trifluoromethoxy)phenylhydrazone (FCCP)
138
. The FCCP treatment led to TMRM fluorescence
loss in all MCF-7 cells (Fig. 2-2A, B). To quantitatively analyze the spatial regulation, we plotted
the radial distribution of TMRM fluorescence within these micropatterns, where the normalized
radial distances of 0 and 1 indicate the approximate center and edge of the micropatterned tumor
island, respectively, and distances greater than 1 represent the surrounding stromal cell areas (Fig.
2-2B). TMRM fluorescence in MCF-7 cells within the micropattern was significantly higher than
the residual TMRM fluorescence post-FCCP treatment at radial distances greater than 0.52 (as
indicated by the red point on the no-treatment (NTX) curve in Fig. 2-2B).
38
Figure 2-2. Differential regulation of mitochondrial membrane potential (ΔΨm) and mass at the
interface vs. center in the µTSA. (A) ΔΨm levels assessed by TMRM fluorescence on Day 4 MCF-
7-BMSC µTSA before and after addition of 20μM FCCP uncoupler. Scale bar: 500 μm; (B)
Representative radial distribution of TMRM fluorescence in a MCF-7-BMSC µTSA on Day 4
before and after FCCP treatment. The red dot (r=0.52) on the no-treatment (NTX) curve indicates
that the TMRM fluorescence to the right of the dot is significantly higher than that of all FCCP-
treated samples; from N=6 independent experiments (p<0.05, Welch’s t-test); (C) µTSA stained
for mitochondrial mass with anti-TOM20. Right panels: confocal scans of the edge and the center.
Green: TOM20; purple: vimentin; blue: DAPI. Scale bars: 500μm in widefield (left) and 25μm in
39
confocal (right). (D) Fold difference in TMRM and TOM20 fluorescence between the edge and
the center. N = 3 independent experiments. P-values: ordinary one-way ANOVA.
Next, we investigated whether the differences in ΔΨm were due to a difference in
mitochondrial mass in the cells at the center and interface. On Day 4 of µTSA cell culture, we
fixed and immunostained cells for TOM20, a protein expressed on the mitochondrial outer
membrane and commonly utilized for mitochondrial quantification
188
(Fig. 2-2C). The
fluorescence intensities of TOM20 and TMRM staining at the interface were both normalized
against those at the center of the µTSA. We found that both mitochondrial mass and ΔΨm were
enhanced at the interface compared to the center, but that the enhancement in ΔΨm (~3.1 fold) was
significantly higher than that of the mitochondrial mass (~1.7 fold) (Fig. 2-2D). The results
indicate that the ΔΨm increase at the interface is mainly due to increased proton gradient, and not
to increased mitochondrial mass.
2.3.3 Fluorescence microscopy of metabolic coenzymes reveals enhanced redox activity at the
interface
To examine the differential metabolic functions in the µTSA, we employed a non-invasive,
label-free approach, using microscopic imaging of the naturally occurring auto-fluorescence of the
metabolic coenzymes, nicotinamide adenine dinucleotide (NADH) and its phosphate ester
NADPH, designated NAD(P)H as their fluorescence cannot be distinguished
189,190
(Fig. 2-3A, B),
as well as flavin adenine dinucleotide (FAD) (Fig. 2-3C, D). The fluorescence of FAD and
NAD(P)H was simultaneously acquired through 540/50 nm and 460/80 nm emission filters with a
single two-photon excitation laser light source at 780 nm. The fluorescence signals were then
40
processed to obtain the optical redox ratio (ORR), which is indicative of OXPHOS metabolism
191
,
and is defined as the ratio of FAD fluorescence over the sum of signal intensity from FAD and
NAD(P)H (Fig. 2-3E). The images were further segmented based on FAD intensity for
mitochondrial regions
192
at the subcellular level to specifically measure the ORR in mitochondria
in individual cells (Fig. 2-3F). We observed altered energy metabolism in the mitochondria of
interfacial cancer cells indicated by their higher ORR when compared to cells at the center of the
µTSA (Fig. 2-3F). The results are consistent with the RNA-seq data showing spatially elevated
mitochondrial OXPHOS at the µTSA tumor-stromal interface.
Figure 2-3. Redox imaging of cancer cells in a µTSA. Representative images of (A) NAD(P)H
fluorescence; (C) FAD fluorescence; and (E) optical redox ratio (defined as
41
FAD/(FAD+NAD(P)H)) at the center and edge of the µTSA on Day 4. Scale bar: 25 μm.
Quantification of (B) NAD(P)H and (D) FAD fluorescence intensities, and (F) the optical redox
ratio at the single-cell level from the center (blue dots) or the edge (red dots) areas within the
µTSA. Mitochondrial regions are segmented from FAD images and applied to the optical redox
ratio images (green regions in C & E). Color scale is to the right of each image. (Representative
dataset from N = 4 independent experiments; p-values: Welch’s t-test.)
2.3.4 ΔΨm spatial profile correlates with YAP/TAZ nuclear localization in the micropatterns
We then aimed to investigate the basis for the spatial differences in ΔΨm within the
micropatterns. We had observed a significant upregulation of actin-cytoskeleton related genes at
the tumor-stromal interface (Fig. 2-1C), which led us to hypothesize that the spatial patterns of
ΔΨm in the µTSA result from the differential cellular constraints experienced by the epithelial
cancer cells within the tumor island. Cells at the interface are weakly confined by stromal cells,
whereas those closer to the center undergo heavier constraints from the neighboring epithelial
cells. To test this hypothesis, in addition to the regular MCF-7-BMSC µTSA, we micropatterned
MCF-7 islands surrounded by a thin layer of micro-contact printed polydimethylsiloxane (PDMS)
to impose complete physical constraints on the island, as well as MCF-7 islands without stromal
cells or PDMS, which lack epithelial or stromal constraints (Fig. 2-4A). Using TMRM staining,
we found distinct differences in the spatial distribution of ΔΨm within the MCF-7 islands in the
three micropatterns (Fig. 2-4B). MCF-7-BMSC µTSA showed higher ΔΨm in MCF-7 cells closer
to the interface, consistent with results in Fig. 2-2. Constraining the tumor island with PDMS led
to an almost complete loss of high ΔΨm at the micropattern edge. However, when the
42
micropatterned tumor island grew free of any physical constraints, a wider band of cells with high
ΔΨm was observed at the periphery of the micropattern (Fig.2-4B).
To get a better molecular understanding of the ΔΨm regulation, we immunostained the
micropatterns for the transcriptional regulators, YAP/TAZ, which can translocate from the cytosol
to the nucleus, reflecting contact-mediated Hippo-signaling
193,194
or mechanical cues experienced
by the cells
195,196
. Figure 2-4B shows representative images of subcellular localization of
YAP/TAZ at the center and edge of the micropatterns. The percentage of cells with nuclear
YAP/TAZ in these regions was calculated (Fig. 2-4C). We found that MCF-7 cells at the edges of
the co-culture or of the monoculture had higher levels of nuclear YAP/TAZ when compared to
their respective centers. When surrounded by PDMS, YAP/TAZ localization across the whole
micropattern became predominantly cytoplasmic, as seen at the center of the mono- and co-
cultures (Fig. 2-4B, C). Interestingly, the percentage of nuclear YAP/TAZ in MCF-7 cells at the
edge of co-cultures was intermediate and significantly different from those at the edges of the
monoculture and PDMS-confined controls, suggesting that stromal cells impose an intermediate
level of physical constraints. Next, we plotted nuclear YAP/TAZ vs. TMRM fluorescence from
the monoculture and co-culture conditions. We picked three locations (edge, center, and
intermediate) from the monoculture and two locations (edge and center) from the co-culture based
on the visible differences in TMRM intensities among these regions. Interestingly, we found a
strong positive correlation between the two parameters (p=0.032, R
2
=0.8279, Fig. 2-4D).
Together, the results suggest that the spatial distribution of ΔΨm of cancer cells in micropatterns
stems from the differential physical constraints imposed on cells.
43
Figure 2-4. Correlation of ΔΨm and YAP/TAZ nuclear translocation in micropatterns. (A)
Schematics of the three micropattern cultures used in this experiment; (B) TMRM staining of ΔΨm
(scale bar: 500 μm) and YAP/TAZ immunostaining (scale bar: 25 μm) in the three micropatterns
on Day 4; (C) Quantification of cancer cells with nuclear YAP/TAZ localization at the center and
edge of the µTSA (n.s.: not significant; ****: p < 0.0001 by ordinary one-way ANOVA);
Representative dataset from N = 3 independent experiments. (D) Liner regression of YAP/TAZ
nuclear localization and TMRM fluorescence in cancer cells in the monoculture and co-culture
µTSA. Three locations (center, edge and approximately 700 μm away from the edge) were taken
from the monoculture µTSA and two locations (center and edge) from the co-culture.
44
(Representative dataset, N = 2 independent experiments; p-value: zero-slope hypothesis in linear
regression).
2.3.5 ΔΨm distribution and YAP/TAZ nuclear translocation in µTSA are modulated by stromal
density
We noticed that the width of the cancer cell region with high ΔΨm at the tumor-stromal
interface of the µTSA co-culture was variable within the same micropattern, as illustrated in Fig.
2-4B. A similar non-uniformity was also seen with nuclear translocation of YAP/TAZ in these co-
cultures, as evidenced by the large standard deviation in Fig. 2-4C. As YAP/TAZ is involved in
contact-mediated growth regulation, we hypothesized that these variations may be due to
differences in local cancer cell densities as a result of differential stromal constraints. To test this
hypothesis, we investigated whether the ΔΨm profile, YAP/TAZ nuclear translocation, and
interfacial cancer cell density could be altered by the initial seeding density of the stromal cells.
MCF-7 cells were micropatterned with three stromal densities (221, 442 or 884 cells/mm
2
,
corresponding to 25k, 50k, and 100k initial seeding numbers in Fig. 2-5A~C), and stained for ΔΨm
with TMRM and YAP/TAZ by immunostaining on Day 4. Consistent with our hypothesis, we
found that the interfacial region of cancer cells with high ΔΨm was much wider in the micropatterns
with a lower stromal density than in those with higher densities (Fig. 2-5A). As controls, the open-
edge and PDMS-confined mono-cultures showed the widest and narrowest regions of cancer cells
with high ΔΨm, consistent with those in Fig. 2-4. To quantitatively compare the ΔΨm profiles, we
analyzed the radial distribution of TMRM fluorescence in the micropatterns. In the open-edge
micropatterns, the cancer cell region with high ΔΨm was wide, with ΔΨm peaking rapidly at a
normalized radial distance as low as 0.67 (the edge is at r=1). In the µTSA co-cultures, however,
45
the width of the high ΔΨm region was narrower and the peak closer to the interface (r=0.85, 0.91,
0.92 as stromal cell seeding density increased) (Fig. 2-5B). In the PDMS-confined micropatterns,
ΔΨm was uniformly low throughout the tumor island, except at the very edge, where cells clump
along the PDMS barrier, leading to a slight increase in observed TMRM fluorescence. Strikingly,
YAP/TAZ nuclear localization followed a similar trend as the ΔΨm. The increased stromal density
led to lower nuclear YAP/TAZ in the interfacial cancer cells (Fig. 2-5C). Noticeably, under the
initial stromal density of 50k per well (442 cells/mm
2
), interfacial cancer cells still had
significantly higher nuclear YAP/TAZ than those in the center. However, such difference was
abrogated with higher stromal seeding density. Cells at the center of all the micropatterns showed
a uniformly low nuclear YAP/TAZ localization (Fig. 2-5C).
Next, we quantitatively analyzed the impact of stromal densities on cancer cell growth and
size, and their relationship with the ΔΨm profile. While cancer cells are generally considered not
to be subjected to growth arrest from contact inhibition
2
, in our µTSA model, the area of the tumor
island with higher stromal density appeared smaller than those seeded with lower stromal density
(Fig. 2-5A,D), suggesting growth retardation and/or size restriction due to the physical constraints
imposed by stromal cells. To simplify the analysis, we initially modeled cell growth in the
micropatterns by measuring the size of the tumor islands. Cell growth is typically described as an
exponential model, 𝑌 = 𝑌 0
𝑒 𝑘𝑡
, where 𝑌 0
is the initial population size, 𝑘 is the growth rate, and 𝑡 is
the growth time
197
. To incorporate the impact of stromal density, we modified the growth curve
to:
𝑌 = 𝑌 0
𝑒 𝑘 (𝑠 )𝑡 (1)
where 𝑘 (𝑠 ) becomes a function of stromal density 𝑠 . Interestingly, using curve fitting, we found a
simple relationship of exponential decay between 𝑌 and 𝑠 (R
2
=0.9599) (Fig. 2-5D). With growth
46
time 𝑡 being a fixed value (Day 4), the result implies that 𝑘 (𝑠 ) is a linear function of 𝑠 with a
negative slope:
𝑘 (𝑠 ) = −𝑘 ′
𝑠 (2)
Here, 𝑘 ′ is a rate constant representing the growth restricting effect from the stromal confinement.
Interestingly, we found that the area of cancer cells with high ΔΨm (normalized by the total area
of the tumor island) also had a similar relationship with stromal density, with a high coefficient of
determination (R
2
=0.9865) (Fig. 2-5E), suggesting a direct role of physical confinement on ΔΨm
distribution. We further investigated whether such regulation is mediated through the cancer cell
size at the tumor-stromal interface. We plotted cancer cell densities at the center and interface of
the µTSA against the initial stromal seeding densities. Indeed, higher stromal seeding density
correlated with higher cancer cell density (thus smaller cancer cell sizes) at the edges of the
micropatterns (Fig. 2-5F). Lastly, a strong negative correlation existed between the density of
cancer cells and the normalized area of cancer cells with high ΔΨm (Fig. 2-5G). These results
suggest that stromal confinement controls the spatial distribution of ΔΨm in cancer cells by
regulating their growth and size.
47
48
Figure 2-5. Regulation of cancer cell ΔΨm by stromal density. (A) TMRM fluorescence in a µTSA
with BMSC seeding densities varying from 25,000, 50,000, to 100,000 cells per micropattern.
Monocultures without or with PDMS constraint were used as controls. Scale bars: 500μm; (B)
Normalized radial distribution of TMRM fluorescence in micropatterns; (C) Percentage of
YAP/TAZ nuclear localization in cancer cells at the edge and center of the micropatterns (n.s.: not
significant, *p < 0.05, ****p < 0.0001 by ordinary one-way ANOVA); (D) Areas of cancer islands
as a function of the initial stromal density on Day 4. Non-liner regression: single-exponential
decay; (E) TMRM peak area at the interface normalized to total cancer area of the respective
micropattern as a function of initial stromal density. Non-linear regression: single-exponential
decay; (F) Cancer cell densities at the center (red curve) and edge (black curve) of micropatterns.
(*p < 0.05, ***p < 0.001, by Kruskal-Wallis test); (G) Normalized TMRM peak area as a linear
function of cancer cell density at the interface. Linear regression R
2
= 0.9957. Representative
dataset shown from N = 3 independent experiments for MCF-7, MCF-7+50kBMSC, and MCF-
7+PDMS micro-patterns, N = 1 for MCF-7+25kBMSC and MCF-7+100kBMSC micropatterns.
2.3.6 Inhibiting the Rho-associated protein kinase and actin polymerization leads to a loss of
ΔΨm spatial distribution
Rho-associated protein kinase (ROCK) and the actin cytoskeleton are upstream of
YAP/TAZ and are intricately involved in both contact-mediated Hippo-YAP signaling and
contact-independent mechano-signaling through YAP/TAZ
195,198
. To investigate the involvement
of ROCK and actin cytoskeleton in the spatial distribution of ΔΨm, we used two chemical
inhibitors, Y-27632 and Latrunculin A (LatA), both of which relax cellular actin tension and
inhibit mechanotransduction
199,200
, and the open-edge monoculture (Fig. 2-4A, right), which
49
eliminates the potential influence of stromal factors other than the physical confinement, while
retaining the spatial distribution of ΔΨm (Fig. 2-4B, right). LatA directly inhibits actin
polymerization, and Y-27632 decreases actin tension by phosphorylating myosin light chain and
activating myosin II
201
. Both Y-27632 and LatA inhibited regular actin polymerization in MCF-7
cells at the center and edge of the micropatterns (Supplementary Fig. 2-S1). At the center of the
micropatterns, the MCF-7 cells originally had actin localization at the cell junctions, which became
discontinuous upon treatment with Y-27632 and LatA. At the edge of the micropattern, those
treated with Y27632 acquired thin lamellar extensions and showed a decrease in actin stress fibers,
while those treated with LatA demonstrated an increased abundance of punctate actin clusters
(Supplementary Fig. 2-S1), as reported elsewhere
196,202
.
As shown in Figure 2-6A & B, both Y-27632 and LatA increased the ΔΨm (TMRM) in
the cancer cells at the center of the micropatterns over a 4-hour period. The increase started as
early as 30 min after treatment, rose rapidly between 30 min and 1 hour, and became slower from
1 to 4 hours (Fig. 2-6C, D). Notably, the two drugs had distinct effects on cancer cells. Y-27632
increased the ΔΨm only at the center of the micropatterns (Fig. 2-6C), while LatA treatment
significantly enhanced the ΔΨm both at the edge and center of the micropatterns at all the time
points (Fig. 2-6D). These results demonstrate a dependence of ΔΨm regulation on actin
cytoskeleton
203
and Rho-ROCK signaling
204
, and on cell location within the micropatterns.
50
Figure 2-6. Loss of ΔΨm heterogeneity by inhibition of mechanotransduction. (A, B) TMRM
fluorescence in monoculture µTSA on Day 4 treated with inhibitors of mechanotransduction: Y-
27632 (50μM, ROCK inhibitor) and Latrunculin A (LatA, 0.5μM, actin polymerization inhibitor)
from 30 to 240 minutes. Scale bar: 500μm; (C, D) Changes of TMRM fluorescence in cancer cells
at the center and edge as a function of treatment time (Representative dataset from N=3
independent experiments; *p < 0.05, ***p < 0.001, ****p < 0.0001, by ordinary one-way
ANOVA).
51
Figure 2-7. Correlation of ΔΨm with metastatic potential in vivo. (A) MCF-7 cells and (B) MDA-
MB-231 cells, both transduced with GFP/luciferase, were sorted into a ΔΨm-high and -low
subpopulations for tail-vein injection into NSG mice; Quantification of the metastatic burden in
the lungs of mice injected with (C) MCF-7 cells at week 5 and (D) MDA-MB-231 cells at week
4, via ex vivo quantification of luciferase activity in tissue lysate (p-values: Mann-Whitney test).
2.3.7 ΔΨm level correlates with metastatic potential in vivo
We next sought to investigate the in vivo significance of the ΔΨm heterogeneity. RNA-seq
data suggested that cancer cells at the tumor-stromal interface have higher ΔΨm and elevated
expression of a gene set related to metastasis (Fig. 2-1C). MCF-7 cells are known to be weakly
52
metastatic
205,206
. Therefore, we used both MCF-7 and MDA-MB-231 cells, the latter of which is a
metastatic breast cancer cell line, to determine if any correlation exists between ΔΨm and
metastatic potential in vivo. MCF-7 and MDA-MB-231 cells were sorted into subpopulations with
high and low-ΔΨm (top and bottom 25%, Fig. 2-7A,B), and then separately inoculated through tail
vein injections into immune-deficient NSG (NOD.Cg-Prkdc
scid
Il2rg
tm1Wjl
/SzJ) mice (the mice for
MCF-7 cells were pre-implanted with estrogen pellets; see Methods). Both MCF-7 and MDA-
MB-231 cells had been previously transduced with a GFP/firefly luciferase-reporter construct for
downstream ex vivo measurement
175,176
. Metastases were allowed to develop for four (for MDA-
MB-231) or five (for MCF-7) weeks post injection. At the end, mice were sacrificed, their lungs
were snap-frozen, crushed, and lysed, and the luciferase signal per tissue weight was measured to
assess metastatic burden
155
. In mice injected with MCF-7 cells, the luciferase signal from the lung
tissues was low compared to those with MDA-MB-231 cells (Fig. 7C,D), and there was no
significant difference between the lung metastatic burden between mice injected with ΔΨm-high
and those with ΔΨm-low MCF-7 cells (Fig. 2-7C). Strikingly, while MDA-MB-231 cells are
known to be highly aggressive, and had a tighter distribution of ΔΨm than MCF-7 cells (Fig. 2-
7B), mice injected with the ΔΨm-high MDA-MB-231 cells had a significantly higher metastatic
burden than those injected with ΔΨm-low cells (p=0.03, Mann-Whitney test, Fig. 2-7D).
53
Figure 2-S1. Fluorescence images showing F-actin staining in MCF-7 cells at the centers and
edges of day 4 micro-patterns, after 4 hours of drug treatment with 50μM Y-27632 or 0.5μM
Latrunculin A, along with their respective no treatment controls. Representative images shown
from three images per condition, scale bar = 25μm.
Figure 2-S2. Quantification of nuclear YAP/TAZ in MCF-7 cells at the centers and edges of day
4 micropatterns, following 12 hours of drug treatment with 50μM Y-27632 or 0.5μM Latrunculin
A, along with their respective no treatment controls. Three fluorescence images per region were
quantified, p < 0.0001 in an ordinary One-Way ANOVA.
54
2.4 Discussion
TME-mediated metabolic reprogramming plays an important role in sustaining tumor
growth and progression
147,158,207-213
. In this study by using a micropatterned co-culture model
mimicking the morphological features of early breast tumors
155,156
, we show that mitochondrial
and metabolic phenotypes can be spatially regulated by surrounding stromal cells under the
architectural context of tumor-stromal interactions. Specifically, mitochondrial mass, ΔΨm and
optical redox ratio (ORR) are enhanced near the tumor-stromal interface as compared to the center
of the tumor nest. Importantly, ΔΨm, a mitochondrial feature previously linked to cancer
invasiveness
13
, is directly controlled by the physical confinement imposed by the surrounding
stromal cells, which correlates with YAP/TAZ nuclear localization. We further demonstrated that
the spatial regulation of ΔΨm by physical confinement is dependent on ROCK signaling and actin
polymerization, and that high ΔΨm correlates with increased metastatic potential in vivo. To the
best of our knowledge, such relationships have not been previously reported in cancer cells. Our
results suggest that cancer cells can perceive physical cues from stromal cells, which may
potentially regulate their metastatic behavior. As this study was done with breast cancer cells, more
studies are needed to further examine the signaling pathways underlying the physical regulation
of mitochondrial heterogeneity, and determine the applicability of these findings to other cancer
and stromal cell types. Notably, some other physical cues within the TME, such as oxygen
gradients and ECM stiffness, are also tightly associated with the tumor architecture
214,215
in early
cancer progression, and have been shown to alter cancer cell metabolism
216-218
. Additional studies
are envisioned to elucidate the contribution of these components to the heterogeneous
mitochondrial phenotypes.
55
In the current study, we uncovered that physical confinement from stromal cells controls
the local density/size of the cancer cells and their ΔΨm level (Fig. 5). Interestingly, it has been
reported that mitochondrial content/mass scales linearly with cell size in HeLa cells
219
, human
umbilical vein endothelial cells (HUVECs)
220
, and budding yeast
221
. More recently, Miettinen and
Bjorklund showed that ΔΨm and OXPHOS follow a non-linear, bell-shaped relationship with cell
size, and those with an ‘optimal’ cell size at the highest ΔΨm seem to have decreased apoptosis
and enhanced cellular proliferation
220
. Our observed co-upregulation of mitochondrial mass, ΔΨm
and OXPHOS in the µTSA thus support the cell size theory and suggest a survival/proliferative
advantage of the interfacial cancer cells in the µTSA. More intriguingly, our study further revealed
an adaptive response of cancer cell size and ΔΨm to the physical confinement by stromal cells in
TME. As stromal density increases, the density of cancer cells at the interface becomes higher;
however, it plateaus at a density that is still much lower than that at the center of the micropattern
(Fig. 2-5F). Nevertheless, while the width of the interfacial cancer cell region with higher ΔΨm
becomes narrower with higher stromal density, the peak fluorescence intensity of the ΔΨm staining
remains similar across all the patterns (except the PDMS confined culture). These results suggest
that the cancer cells near the tumor-stromal interface can accommodate and adjust their sizes to
retain an ‘optimal’ mitochondrial state/functionality, which is absent in the PDMS-confined
culture.
In this study, we also observed preferential localization of YAP/TAZ in the cytoplasm or
nuclei of cells at the center or interface of µTSA, respectively. This observation is in agreement
with previous findings on YAP/TAZ regulation by cell density
198
. Importantly, at the population
level, we further revealed a linear correlation between ΔΨm and YAP/TAZ nuclear translocation
in the micropatterns (Fig. 2-4D). It is unclear, however, whether this represents a causal
56
relationship. It has been shown that YAP overexpression in breast cancer cells leads to increased
mitochondrial mass
222
. On the other hand, PLD6-mediated increase in mitochondrial fusion was
shown to inhibit YAP/TAZ activity through AMPK-mediated YAP/TAZ phosphorylation in
human mammary epithelial cells
223
. These studies and ours suggest a crosstalk between
mitochondrial dynamics and YAP/TAZ activation. Additional studies are needed to further
elucidate the signaling relationship between ΔΨm and YAP/TAZ.
It is noteworthy that nuclear YAP/TAZ has been found to act as relays of mechanical
signals resulting from cell density
198
, ECM stiffness
195
, and intracellular cytoskeletal tension
201
.
Recent studies have shown mitochondrial functions to be actin-dependent. Mitochondrial fusion
and fission rely on dynamic actin polymerization
98
. ΔΨm cannot be maintained in mouse
embryonic fibroblasts when β-actin is knocked out
224
. On the other hand, Furukawa et al. have
recently shown that the cell density dependence of YAP/TAZ localization is mediated through
intracellular actin tension
201
. Cellular actin tension was reduced in the micropatterns using LatA
and Y-27632. The distinct changes in cell morphology and actin localization upon treatment with
LatA and Y-27632 were consistent with previously reported results
196,202
. On the other hand, LatA
and Y-27632 treatments did not lead to differences in nuclear YAP/TAZ, even after increasing
treatment time to 12 hours (Supplementary Fig. 2-S2). The lack of response of YAP/TAZ
localization to both drugs is consistent with a previous report that the dependence of YAP/TAZ
localization on F-actin integrity occurs only in the absence of cell-cell contacts, but not with
confluent epithelial cells
196
. Both these drugs led to an increase in cellular ΔΨm in cancer cells at
the center of the micropatterns (Fig. 2-6), confirming actin stress polymerization is upstream of
ΔΨm regulation. Interestingly, at the edge of micropatterns, LatA treatment increased ΔΨm while
Y-27632 has no effect, indicating a difference in the ΔΨm regulation in cancer cells between the
57
edge and center of the micropattern. A possible explanation is the difference in the activation of
YAP signaling in these two regions. Activated nuclear YAP has been shown to negatively regulate
Rho
225,226
, and Rho activity is required to activate ROCK through its Rho-binding domain
227
. The
difference in YAP/TAZ nuclear translocation between the two spatial regions could thus explain
the difference in sensitivity to ROCK inhibition. The differential outcome with inhibition of
cytoskeletal tension also suggests that cancer cells may experience different mechanical stresses
(tensile vs. compressive) at the interface and center of the µTSA, which have been shown to
associate with tumor architecture in a growing tumor mass
160
. Future studies aiming at measuring
the distribution of stresses in the µTSA are envisioned.
Importantly, while our study points to the physical confinement imposed by the stromal
cells as the main source of spatial ΔΨm distribution, it does not rule out other mechanisms that may
also regulate mitochondrial functions, such as biochemical signaling through cell-cell adhesion
and soluble factors. One such indication comes from the fact that the interfacial cancer cells under
high stromal density do not behave as the PDMS-confined cancer cells (Fig. 2-5A, B).
Interestingly, at the highest stromal seeding density (100K), YAP/TAZ was located in the
cytoplasm both at the tumor-stromal interface and at the center (Fig. 2-5C); however, under the
same conditions, the tumor-stromal interface still retains a thin layer of cancer cells with high ΔΨm.
One potential mechanism of this interface-bulk difference is the metabolic interactions between
tumor and stromal cells, as well as between tumor cells from different micropattern regions. For
instance, both cancer-associated fibroblasts (CAFs)
228
and hypoxic tumor cells
229
have been shown
to undergo glycolysis and to supply lactate to fuel cancer cells that engage in mitochondrial
OXPHOS. Notably, while there is no apparent hypoxia in the µTSA, cancer cells in the center
have upregulated glycolytic gene expression compared to the interfacial cells (Fig. 2-1C). Another
58
potential mechanism is through heterotypic tumor-stromal adhesion or cancer-ECM interactions
at the epithelial-stromal interface. For example, breast cancer cells stimulate CCL5 production
from BMSCs upon tumor-stromal contact
177
, which increases glucose uptake and ATP production
in breast cancer cells
230
. Breast cancer cells can also secrete growth factors to activate nearby
stromal cells
231
, which remodel the physical and biochemical properties of the surrounding ECM
through matrix metalloproteinases
232
and production of collagen
233,234
. Future studies are needed
to determine the exact contributions of these biochemical factors to the heterogeneous
mitochondrial phenotypes in µTSA.
While previous studies have reported association of high ΔΨm in cancer cells with
increased secretion of vascular endothelial growth factor (VEGF) and matrix metalloproteinase-7
(MMP-7) as well as enhanced invasiveness in vitro
13
, the in vivo significance of ΔΨm in cancer
metastasis remained unclear. Our RNA-seq data suggested increased expression of metastasis
related genes in interfacial cancer cells (Fig. 2-1C), leading us to hypothesize that the ΔΨm of
breast cancer cells is associated with their in vivo metastatic potential. One of the challenges of
testing this hypothesis is that the MCF-7 breast cancer cells are only weakly metastatic in vivo
205,206
and MCF-7 mouse models are generally considered to model early-stage breast cancer
235
. We
inoculated mice with MCF-7 cells sorted based on their ΔΨm and assayed the lungs for metastatic
burden five weeks post injections. Not surprisingly, the lung metastatic burden with MCF-7 cells
was low despite of the higher inoculation doses than the MDA-MB-231 injection group. This is in
accordance with a previous report which showed that MCF-7 cells injected through tail vein
injection only resulted in scattered cancer cell presence in the lungs even after nine weeks, as
opposed to large multicellular lung metastases in those injected with MDA-MB-231 cells in just
two weeks post injection
206
. Further, we observed no difference in the lung metastatic burden
59
between the groups injected with ΔΨm -high vs. ΔΨm -low MCF-7 cells (Fig. 2-7C), which could
have resulted from the overall low metastasis forming ability of the MCF-7 cells. In stark contrast,
we found that high ΔΨm MDA-MB-231 cells, when injected into mice, formed significantly more
lung metastases than those injected with low ΔΨm cells. Together, these results suggest that while
the ΔΨm may not be sufficient by itself to promote metastasis, it contributes to the metastasis
formation in later stage cancer progression (as supported by the MDA-MB-231 metastasis model).
Some mitochondrial activities have been previously shown to be involved in the process
of breast cancer metastasis in vivo. PGC1-α, a master regulator of mitochondrial biogenesis, was
found to be essential for lung metastasis formation in mice
151
. We demonstrated, for the first time,
that breast cancer cells (MDA-MB-231) with higher ΔΨm have greater metastatic potential in vivo
(Fig.2- 7), which underscores the importance of investigating TME-mediated mechanisms
governing mitochondrial heterogeneity. Further studies are also necessary to elucidate the
relationship between mitochondrial biogenesis and ΔΨm, and the regulatory role of ΔΨm in
metastatic cascades, such as persistence in circulation, extravasation, and survival at the metastatic
sites.
In summary, we have shown the impact of tumor-stromal interactions in inducing spatial
heterogeneity of mitochondrial activities in the TME, with implications on the metastatic potential
of cancer cells in vivo. We demonstrated that the differential ΔΨm of cancer cells is mainly
controlled by physical confinement from stromal cells, correlates with YAP/TAZ nuclear
translocation, and is mediated through the actin cytoskeleton. Our study provides new insights into
the role of TME in the regulation of mitochondrial heterogeneity and cancer metastasis.
60
Chapter 3: E-Cadherin Regulates Mitochondrial Membrane
Potential in Cancer Cells
3.1 Rationale
Epithelial cancer cells have higher mitochondrial membrane potential ( m) than their
normal counterpart cells
10
, which has been associated with cancer stem cell features, increased
secretion of angiogenic factor and matrix metalloproteinase, as well as higher invasiveness in vitro
13,22,23,236
. We have previously reported in a xenograft metastatic breast cancer model in mice that
cancer cells with higher m result in a greater lung metastatic burden than those with low m
237
. Together these results highlight the biological significance of m in cancer cell. However,
the mechanisms by which it is differentially regulated in situ remain unclear.
The tumor microenvironment (TME) is a complex amalgamation of many types of cues,
including different cell types like fibroblasts and immune cells
238
, biochemical cues from cellular
metabolism/hypoxia and cell-type specific secretions or interactions
239-241
, and physical cues such
as solid stresses and matrix stiffness from tumor growth and extracellular matrix remodeling
116
.
Among those, stromal cells have been found to fuel mitochondrial metabolism in cancer cells
through metabolic coupling
6,242
, while hypoxia-driven induction of transcription factors like
PGC1- increases mitochondrial biogenesis in cancer cells
151
. Importantly, recent studies show
an emerging role of mechanical cues from the TME such as ECM stiffness in influencing cancer
cell metabolism through mechanotransduction, adhesion receptor signaling, and cytoskeletal
reorganization
217
. We have lately reported a spatial distribution pattern of m in cancer cells
associated with physical confinement cues from the surrounding stromal cells using a
micropatterning platform, the micropatterned tumor-stromal assay ( TSA)
155,237
. We showed that
61
the physical confinement from TME downregulates m in cancer cells, while the m of those
without confinement remains high
237
. Yet, the exact mechanisms by which the physical
confinement regulates cancer cell m remain to be determined.
In cancer cells, physical confinement has been found to induce changes in cell adhesion
and cytoskeletal rearrangements, which alter their invasiveness and metastatic potential
243,244
. In
particular, loss of E-cadherin, which forms a core component of intercellular adherens junctions
(AJs)
245
, is associated with increased migration and invasiveness in vitro and exacerbated lung
metastases in vivo
246
. On the other hand, activating E-cadherin adhesions inhibits tumor
metastases and decreases numbers of circulating tumor cells (CTCs) in the blood
247
. Lately, it was
shown that E-cadherin plays a role in limiting oxidative stress and reactive oxygen species (ROS)-
mediated apoptosis in cancer dissemination
248
. Whether E-cadherin regulates pathways directly
affecting mitochondrial activity remains to be investigated, which could provide novel targets for
cancer therapeutics.
In the present study, we investigated whether physical confinement cues can induce
spatially regulated cell adhesion and how these in turn regulate m level and its spatial
distribution within TME. We show that pathways related to E-cadherin-mediated AJs are
differentially regulated at the edge vs. center of the tumor model and that E-cadherin expression
correlates with m spatial distribution. We further demonstrated that disrupting AJs rescues the
m level in confined cancer cells with lower m, while overexpressing E-cadherin decreases
the m level at the micropattern centers. Our work thus provides a novel insight into the potential
role of E-cadherin mediated adhesions in regulating m in cancer cells.
62
3.2 Materials and Methods
3.2.1 Cell culture and micropatterning
MCF-7 and MDA-MB-231 cells were obtained from American Type Culture Collection
(ATCC) cultured as described previously
237
in Dulbecco’s Modified Eagle Medium (DMEM)
with 10% Fetal Bovine Serum (FBS), 100 U/ml penicillin and 100 g/ml streptomycin (P/S). For
the open edge unconfined micropatterns, an Epilog laser engraver was used to cut 2 mm diameter
holes within a 10 mm diameter circular pattern in a 250 m thick polydimethylsiloxane (PDMS)
sheet. These PDMS stencils were aligned on collagen coated coverslips
237
, treated with 0.2 %
Pluronic F-127 (Sigma) and rinsed with PBS prior to cell seeding. Either of the following two
methods were used for cell seeding. For cell seeding, 300,000 cancer cells were seeded per well in
1 ml of DMEM per well of a 24-well plate, with each well containing a PDMS stencil aligned onto
a collagen coated coverslip. The well-plate was centrifuged at 200 x g for 5 minutes followed by
the slowest deceleration setting. The cells were then incubated for 4 – 5 hours, before micropatterns
were rinsed with DMEM and placed in a fresh well with new DMEM and incubated for 4 days
prior to use in experiments. For confined cancer cell micropatterns, master molds (array of 500
μm diameter islands) were designed using AutoCAD (Autodesk) and fabricated through
photolithography on silicon wafers using SU8 photoresist (100 m thick). Sylgard 184
polydimethylsiloxane (PDMS, 10:1 base: curing agent) (Dow Corning) was poured on the wafers,
degassed and cured overnight at 65
0
C following which PDMS stamps were cut out using a 10 mm
biopsy punch. Freshly mixed liquid PDMS (10:1 mix ratio) was spun coated onto 40 mm x 24 mm
rectangular glass coverslips. These were incubated at room temperature for 45 minutes, then the
PDMS stamps were dipped into the spun coated liquid PDMS and printed onto collagen coated
63
coverslips. The coverslips were incubated overnight at room temperature to cure the PDMS,
treated with 0.1% Pluronic, and rinsed with PBS prior to cell seeding (previously described seeding
method; 300,000 MCF-7 cells per well). Confined cell micropatterns were cultured for 4 days to
allow the cadherin-dominant micropatterns to form prior to experiments.
3.2.2 Generation of E-cadherin-GFP expressing and E-cadherin knockout cell lines
Plasmid DNA encoding E-cadherin-GFP was obtained from Addgene (plasmid # 28009
deposited by Jennifer Stow; http://n2t.net/addgene:28009; RRID:Addgene_28009)
249
. Plasmid
DNA was amplified with DH5α (Thermo Fisher) and isolated using the QIAprep Spin Miniprep
Kit (Qiagen) according to manufacturer’s instructions, and the sequence was confirmed by Sanger
sequencing with CMV-F, EGFP-N and BGH-rev primers at GENEWIZ. 200,000 MDA-MB-231
cells and 150,000 MCF-7 cells were seeded in 6-well plates and after overnight incubation,
transfected with E-cadherin-GFP plasmid DNA using the Effectene Transfection Reagent
(Qiagen) according to manufacturer’s instructions (0.4 μg plasmid DNA per transfection). Cell
culture media was changed 24 and 48 hours post transfection, and cells were then passaged 1:5 in
antibiotic selection media (DMEM, 10% FBS, 0.5 mg/ml geneticin, no P/S). Antibiotic selection
was maintained till there were no cell colonies growing in the non-transfected control wells (7-10
days). Transfected cells were then expanded and FACS sorted for GFP positive cells. Clustered
regularly interspaced short palindromic repeats (CRISPR) technology was used to generate E-
cadherin knockout (KO) MCF-7 cells. Briefly, 150,000 MCF-7 cells were seeded in a 6-well plate
and allowed to adhere overnight. The next day, cells were trans-fected with 0.4 µg of E-cadherin
CRISPR/Cas9 KO plasmids (sc-400031, which encode E-cadherin-specific 20 nt guide RNA
sequences, SpCas9, and GFP reporter) using Effectene Transfection Reagent (Qiagen). Cell
64
culture media was changed 24 hours and 48 hours post transfection. E-cadherin KO cells were
then harvested and FACS sorted by positive GFP fluorescence (transiently expressed by the
transfected cells). Sorted KO cells were expanded for subsequent studies.
3.2.3 Mitochondrial membrane potential staining and imaging
Micropatterns were incubated in extracellular imaging buffer (130 mM sodium chloride, 5
mM potassium chloride, 1.5 mM calcium chloride, 1 mM magnesium chloride, 25 mM 4-(2-
hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 1 mg/ml BSA, 5 mM glucose, pH
adjusted to 7.4) with 10 nM TMRM (Life Technologies) for 45 minutes, and imaged in the same
dye-containing buffer using a Nikon Eclipse Ti inverted microscope, using a Nikon Plan Fluor 10x
objective with a numerical aperture (NA) of 0.30 (for unconfined micropatterns) or a Nikon Plan
Apo 20x objective with 0.75 NA (for confined micropatterns). A Nikon C2 confocal microscope
(Nikon Plan Apo 60x oil immersion objective, 1.40 NA) was used for confocal imaging.
3.2.4 Drug treatment and immunostaining
After 4 days of culture, micropatterns were treated with 1 mM or 10 mM 1,4-Dithiothreitol
(DTT, Millipore Sigma), live imaged for m with tetramethylrhodamine methyl ester (TMRM,
Life Technologies), following which they were fixed in 4 % paraformaldehyde (PFA) for 15
minutes, rinsed and stored in PBS at 4
o
C till they were immunostained. For antibody mediated
inhibition of E-cadherin adhesions, unconfined MCF-7 micropatterns were treated at day 4 with
50 g/ml anti-E-cadherin antibody (CD324 Monoclonal Antibody, clone DECMA-1,
eBioscience
TM
) for 3 hours in extracellular imaging buffer, and monitored for m changes under
10 nM TMRM. For immunostaining, micropatterned cells on coverslips were permeabilized with
65
0.1 % Triton X-100 for 10 minutes, and blocked in 4 % bovine serum albumin (BSA) for 2 hours.
To validate CRISPR knockout of E-cadherin in MCF-7 cells, DECMA-1 (5 µg/mL) was used for
E-cadherin immunostaining. Samples were then incubated in primary antibody (E-cadherin 24E10,
Cell Signaling, 1:200, TOM20 sc-17764, Santa Cruz Biotechnology, 1:100) diluted in 4 % BSA
for 2 hours, rinsed with PBS 3 times with 5 minutes for each rinse. Following incubation in
secondary antibody diluted in 4 % BSA for 1 hour. After the immunostaining, samples were
further rinsed with PBS 3 times for 5 minutes per rinse, and mounted onto a glass slide using
fluoro-gel II mounting medium (Electron Microscopy Sciences) before microscopic imaging.
3.2.5 Image analysis and quantification
Fluorescence images were analyzed on ImageJ (NIH). For radial distribution plots, a region
of interest (ROI) was drawn around the micropatterned island, and a custom macro was used to
extract pixel coordinates (normalized to the centroid) and pixel intensities within the selected ROI.
A custom MATLAB code was then used to convert the Cartesian coordinates to normalized polar
coordinates, where the radial distances of 0 and 1 represent the center and edge of micropatterns
respectively. For region-based (center vs. edge) quantification of fluorescence, a custom ImageJ
macro was used to select 5 ROIs each at these locations and calculate the average fluorescence
(representative center and edge ROIs are shown in Fig. 3-3a). The ITCN plugin on ImageJ was
used for cell density quantifications. For the quantification of cell-cell contact index, outlines of
individual cells were traced on ImageJ, and the index was calculated by dividing the overlapped
area (AND) by the total combined contour area (OR).
3.2.6 Statistical analysis
66
Data were presented in mean ± S.D. (standard deviation). All statistical analyses (unpaired
t-tests, one-way ANOVA and 2-way ANOVA) were performed on GraphPad Prism and the
resulting p-values are indicated for each figure. N.s.: not significant (p > 0.05), *: p < 0.05, **:
p<0.01, ***: p<0.001, ****: p<0.0001.
3.3 Results
3.3.1 Adherens junctions (AJs) are downregulated at the tumor-stromal interface in a
micropatterned tumor model
We have previously established a micropatterning platform, the micropatterned tumor-
stromal assay ( TSA), to recapitulate tumor-stromal interactions
155
, and demonstrated a spatial
gradient of m in MCF-7 breast cancer cells within the tumor island surrounded by bone marrow
stromal cells (BMSCs)
237
. Within µTSA, MCF-7 cells in regions near the center of the
micropattern had lower m (visualized by the red TMRM staining) than those closer to the tumor-
stromal interface (Fig. 3-1a). Quantitative analysis showed a close to 3-fold difference in m
level between the two regions (Fig. 3-1b). We extracted MCF-7 cells from the center and interface
of the tumor islands using laser capture microdissection (LCM), and performed RNA sequencing
to examine the differential regulation of gene expression between the two regions
237
. Gene set
enrichment analysis (GSEA)
170,171
revealed a significant negative enrichment of pathways related
to adherens junctions (AJs) in MCF-7 cells at the interface relative to the center (Fig 3-1c),
suggesting a spatial distribution of differential cell adhesions (mediated by AJs) within the tumor
island that negatively correlates with m spatial distribution. Since confinement cues were shown
67
to induce changes in cancer cell adhesion in vitro
244
, we hypothesized that the physical
confinement cues induce m changes by regulating the level of AJs in cancer cell adhesion.
Figure 3-1: Spatial distribution of m of MCF-7 cells in micropatterned tumor model associated
with regulation of cell adhesion. (a) Representative image showing TMRM fluorescence of a day
4 MCF-7-BMSC co-culture micropattern and (b) the corresponding normalized radial distribution.
(c) Gene set enrichment analysis of MCF-7 cells at the tumor-stromal interface relative to MCF-7
68
cells at the center of the tumor island, following RNA-sequencing of laser capture microdissected
from different locations of the micropattern as described in
237
with a false discovery rate (FDR) <
0.25.
3.3.2 E-cadherin expression correlates with spatial distribution of m within tumor
micropattern
To eliminate the impact of tumor-stromal biochemical signaling
155
, we created a
micropatterned monoculture of MCF-7 cells on collagen coated coverslips (Fig. 3-2a). After 4
days of culture, MCF-7 cells also formed a spatial pattern of m distribution with low m in
the center and high m at the edge (Fig. 3-2b), although the area of cells with higher m was
greater than those in the co-cultured micropatterns (Fig. 3-1a). Since the MCF-7 monoculture
micropatterns retained the center-edge spatial m gradient, we used this model and its fully
confined variant to assess the role of spatial confinement and cell-cell adhesion in regulating m
levels for the rest of the study. We first examined whether there was a differential pattern of AJ
formation within the micropatterns. We immunostained the micropatterns against E-cadherin, a
core component of AJs in epithelial cells
245
. Confocal imaging revealed that MCF-7 cells at the
edges of the micropattern had lower E-cadherin expression with cytoplasmic localization (Fig. 3-
2b). In contrast, MCF-7 cells at the center of micropatterns showed higher E-cadherin expression
with distinct cell membrane localization, forming a cobblestone-like structure characteristic of
epithelial monolayer that is mediated by AJ formation
250
.
To examine the role of E-cadherin in regulating spatial distribution of m within the
micropatterns, we next picked a breast cancer cell line that has lower/negative E-cadherin
expression to test if they form a different m pattern. The Genevestigator database
251
shows that
69
MDA-MB-231, a metastatic breast cancer cell line, has significantly lower E-cadherin expression
than MCF-7 cells in data collected from >40 independent experiments (Fig. 3-2c). When MDA-
MB-231 cells were micropatterned and live imaged for m, they demonstrated a spatial
distribution of m levels distinct from that in MCF-7 micropattern, where the m of MDA-
MB-231 cells was similar/slightly higher at the center than those at the edge of the micropatterns
(Fig. 3-2d, e). E-cadherin immunostaining and confocal imaging of MDA-MB-231 cells in the
micropattern confirmed that E-cadherin expression in these cells was essentially absent at the cell
membrane and displayed similar intracellular characteristics between cells at the edge and center
of the micropattern (Fig. 3-2c). Together, these results suggested a potential role of E-cadherin-
mediated AJ formation in regulating m in cancer cells.
70
Figure 3-2: Correlation of E-cadherin expression with spatial distribution of m. (a) Schematics
of creating an unconfined monoculture micropattern. (b) Widefield low-magnification imaging
showing spatial distribution of TMRM fluorescence and high-magnification confocal imaging
showing E-cadherin localization in day 4 MCF-7 monoculture micropatterns. (c) Average E-
cadherin (CDH1) expression in MCF-7 cells compared to MDA-MB-231 cells from > 40
independent experiments (obtained from Genevestigator database
251
). **** P < 0.0001 in an
unpaired t test. (d) Spatial distribution of TMRM fluorescence and confocal images of E-cadherin
71
staining in day 4 MDA-MB-231 micropatterns. (e) Average normalized radial distribution of
TMRM fluorescence of MCF-7 and MDA-MB-231 day 4 micropatterns (>=3 micropatterns per
condition).
3.3.3 Disrupting AJ formation increases m in MCF-7 micropattern
We next aimed to investigate the effect of disrupting E-cadherin mediated AJs on the
spatial distribution of m in MCF-7 micropatterns. We used 1,4-dithiothreitol (DTT), a reducing
agent that disrupts E-cadherin mediated cell-cell adhesion by cleaving the disulfide bonds in the
extracellular domains of E-cadherin
252
. At a concentration of 10 mM, DTT has been shown to
selectively disrupt AJs in MDCK cells
253
. We treated MCF-7 micropatterns at day 4 with 1 mM
and 10 mM DTT, and observed a significant increase in m in MCF-7 cells at the centers of the
micropatterns compared to the untreated control (Fig. 3-3a, b). On the other hand, in MCF-7 cells
at the edges of the micropattern, only the higher DTT concentration (10 mM) led to a significant
increase in m. Confocal imaging of E-cadherin immunostaining in MCF-7 cells revealed that
the 10 mM DTT treatment significantly decreases the E-cadherin level per cell at the center of the
micropattern (Fig. 3-3c, d). Moreover, we saw a dose-dependent decrease of fluorescence intensity
in E-cadherin at intercellular junctions with DTT treatment, with 10 mM showing a more markedly
decrease than the 1 mM DTT treatment (Fig. 3-3e). Interestingly, we noticed that, while the lower
DTT concentration (1 mM) did not significantly reduce AJ area (Fig. 3-3d), it was sufficient to
increase m in MCF-7 cells at the micropattern center. We thus tested the response time of m
to the DTT treatment using the 1 mM DTT concentration. We created a confined micropattern of
MCF-7 cells with a thin surrounding layer of PDMS (Fig. 3-3f). After 4 days of culture, MCF-7
cells formed a cadherin-dominant micropattern with uniformly high E-cadherin level at cell-cell
72
junctions throughout the tumor island (Fig. 3-3f). As expected, the m of the MCF-7 cells in the
micropattern became very low (Fig. 3-3g), which was similar to that at the center of the open edge
micropatterns. Upon treatment with 1 mM DTT, we observed a significant increase in the m
level as soon as after 2 hours into the treatment (Fig. 3-3g, h). To further validate the impact of
disrupting E-cadherin mediated AJ formation/cell-cell adhesion, we treated MCF-7 micropatterns
with a function-blocking E-cadherin monoclonal antibody, DECMA-1, which has been reported
to disrupt E-cadherin mediated AJs in MCF-7 cells
254
(Fig. 3-3i). Similar to the DTT treatment,
DECMA-1 treatment significantly increased m of cancer cells at the center, but not at the edge
of unconfined micropatterns (Fig. 3i, j). These results suggest that the AJ formation by E-cadherin
in cancer cells negatively regulates the m level in MCF-7 cancer cells.
73
74
Figure 3-3: Disruption of AJs with DTT in MCF-7 micropatterns. (a) TMRM fluorescence of day
4 MCF-7 unconfined micropatterns with and without 1 mM and 10 mM DTT treatment (3 hours).
(b) Quantification of average TMRM fluorescence at the centers and edges of the micropatterns
shown in (A). * P < 0.0332, ** P < 0.0021, ****P < 0.0001 in a 2-way ANOVA. (c) Confocal
imaging showing E-cadherin fluorescence at the centers of day 4 MCF-7 unconfined micropatterns
with and without indicated DTT treatment. (d) Average E-cadherin area per cell in MCF-7 cells
shown in (c). *** P < 0.0002, ****P < 0.0001 in an ordinary one-way ANOVA. (e) Line scans
showing average E-cadherin fluorescence across intercellular cadherin adhesions as shown in
schematic. At least 18 cell pairs were analyzed per condition. (f) E-cadherin staining showing AJ
formation in an MCF-7 micropattern confined with a thin layer of PDMS. (g) TMRM fluorescence
of MCF-7 cells in confined micropatterns before and after 2 hours and 4 hours of 1 mM DTT
treatment. (h) Quantification of TMRM fluorescence in MCF-7 confined micropatterns treated
with 1 mM DTT over a time period of 4 hours. ****P < 0.0001 in an ordinary one-way ANOVA.
(i) TMRM fluorescence of MCF-7 cells in unconfined micropatterns without or with 50 µg/ml
anti-E-cadherin (DECMA-1) treatment for 3 hours. (j) Quantification of ΔΨm at the center and
edge of micropatterns shown in (i). ns: not significant (p>0.05); * P < 0.05 in a 2-way ANOVA.
3.3.4 E-cadherin expression in MDA-MB-231 cells decreases ΔΨm at the micropattern center
We further examined whether re-expression of E-cadherin in MDA-MB-231 cells, which
have low/no E-cadherin expression, would induce a spatial regulation of ΔΨm levels in the
micropattern. We transfected MDA-MB-231 cells with an E-cadherin-GFP construct
249
, and
created open-edge micropatterns with these cells alongside with the wild-type (WT) MDA-MB-
231 cells as control (Fig. 3-4a, bottom). We confirmed the expression of E-cadherin by observing
75
the E-Cadherin-GFP signal in the micropattern, which was higher at the center than the edge (Fig.
3-4b). We also monitored the spatial distribution of ΔΨm in the micropatterns with TMRM live
staining (Fig. 3-4a, top). ΔΨm at the center of micropattern with MDA-MB-231 cells expressing
E-cadherin was lower than that with WT MDA-MB-231 cells. Although we did not observe a
similar edge vs. center pattern of the ΔΨm levels in MDA-MB-231-Ecad-GFP cells as with MCF-
7 cells, the region with downregulated ΔΨm levels (vs. control cells) correlated with the elevated
E-cadherin-GFP signal at the center of micropattern (Fig. 3-4b, c). The expression of E-cadherin
in transfected MDA-MB-231 cells and the center-edge difference were further confirmed with
immunostaining and regional quantification in micropatterns (Fig. 3-4d, bottom; Fig. 3-4e). We
also assessed whether the decrease in ΔΨm at the micropattern center in the E-cadherin expressing
cells was due to a decrease in mitochondrial mass. Immunostaining of these micropatterns against
TOM20, a mitochondrial protein indicative of mitochondrial mass
237
, revealed that there was no
difference in mitochondrial mass at the centers of these micropatterns (Fig. 3-4d, f). Interestingly,
there was significantly lower mitochondrial mass at the edge of micropattern with MDA-MB-231-
Ecad-GFP cells, where no ΔΨm difference was observed, further supporting the notion that
mitochondrial mass did not contribute to the ΔΨm differences.
Our RNA-seq and inhibition experiments pointed to the importance of E-cadherin mediated
cell-cell adhesion (AJs) in ΔΨm regulation. Under high-magnification confocal microscopy,
although we did not observe a completely epithelial morphology in the E-cadherin expressing
MDA-MB-231 cells, these cells did exhibit morphological changes and increased cell-cell contact
through overlaps of cell protrusions and/or cell bodies (Fig. 3-4g). We defined a cell-cell contact
index as the ratio of the overlapped area between two adjacent cells over the total contour area of
the two cells (Fig. 3-4h). We found that there was significantly higher cell-cell overlap/contact in
76
micropatterns with E-cadherin expressing cells than WT MDA-MB-231 cells both at the center
and edge of micropattern, whereas the difference was more significant at the center than the edge
(Fig. 3-4h). We also ruled out the possibility that such changes were caused by differences in cell
density (Fig. 3-4i). Together, these results indicate that re-expression of E-cadherin in MDA-MB-
231 cells lowers the ΔΨm at micropatterns center through E-cadherin mediated cell-cell adhesion.
77
Figure 3-4: Effect of E-cadherin expression on the spatial distribution of ΔΨm in MDA-MB-231
micropatterns. (a) TMRM and E-cadherin-GFP fluorescence of day 4 unconfined micropatterns
of MDA-MB-231 (non-transfected control) and MDA-MB-231 cells transfected with E-cadherin-
78
GFP (widefield imaging). (b) Radial distribution of E-cadherin-GFP in E-cadherin-GFP
transfected vs. non-transfected micropatterns as shown in (A), n = 4 micropatterns per condition.
(c) Radial distribution of ΔΨm of representative micropatterns shown in (a), n = 4 micropatterns
per condition. (d) Immunofluorescence widefield imaging showing TOM20 and E-cadherin
antibody fluorescence in E-cadherin-GFP transfected vs. non-transfected controls. Quantification
of mean fluorescence intensities of (e) E-cadherin and (f) TOM20 immunostaining at the centers
and edges of the immunostained micropatterns shown in (d). (g) Confocal imaging showing E-
cadherin fluorescence at the centers and edges of MDA-MB-231 and MDA-MB-231-ECadherin-
GFP micropatterns. Quantification of (h) cell-cell contact index and (i) cell density at the centers
and edges of the micropatterns. Contact index defined as the ratio of cell-cell overlap area and total
area of the contour. For (b), all data points on the two curves are statistically different from each
other at each radius (p < 0.05 by t-test); for (c), data points under the solid line are statistically
different from each other at each radius (p < 0.05 by t-test); for e, f, h, i: ** P < 0.0021, *** P <
0.0002, **** P < 0.0001 in a 2-way ANOVA.
3.3.5 E-cadherin knockout and overexpression alter m at the center of MCF-7 micropattern
MCF-7 cells express high levels of E-cadherin compared to MDA-MB-231 cells (Fig. 3-
2c). We next investigated whether knocking out or further overexpressing E-cadherin could affect
the m of MCF-7 cells within the micropatterns. To create E-cadherin knockout (KO) cells, we
transfected MCF-7 cells with a commercial E-cadherin CRISPR/Cas9 knockout kit that contains
3 plasmids, with each encoding Cas9 and a specific guide RNA sequence against CDH1 (E-
cadherin) site in the genomic DNA. To validate the knockout, we immunostained the transfected
MCF-7 cells in unconfined micropatterns at day 4 with the DECMA-1 E-cadherin antibody (Fig.
79
3-5a). We observed a marked decrease in E-cadherin immunostaining at the intercellular junctions
in the center of micropatterns formed by the E-cadherin KO cells compared to the WT MCF-7
cells (Fig. 3-5a, top panels). At the micropattern edges, both WT and KO cells expressed minimal
E-cadherin at the cell-cell borders (Fig. 3-5a, lower panels), similar to what was observed in
immunostaining with 24E10 E-cadherin antibody in micropatterns of WT cells (Fig. 3-2b).
To create E-cadherin overexpressing (OE) cells, we transfected MCF-7 cells with the same E-
cadherin-GFP construct
249
used for MDA-MB-231 cells in Fig. 3-4. We confirmed the
overexpression of E-cadherin by both GFP fluorescence and immunostaining (Fig. 3-5b). Notably,
we saw a fraction of WT MCF-7 cells in the E-cadherin OE micropatterns due to incomplete killing
of WT MCF-7 cells in antibiotic selection (Fig. 3-5b, DAPI+GFP- cells). While both WT and E-
cadherin OE cells had high expression of E-cadherin in micropattern center, the E-cadherin OE
cells (GFP+) had visibly higher E-cadherin immunostaining than the WT (GFP-) cells.
Importantly, at the micropattern edge, the E-cadherin OE cells demonstrated AJ formation
indicated by the presence of E-cadherin-GFP and immunostaining signals at the cell-cell boundary.
In contrast, those WT (GFP-) cells had negligible E-cadherin immunostaining in the same region
(Fig. 3-5b, lower panels).
We then measured the differences of m spatial distribution in unconfined micropatterns with
WT, E-cadherin KO, and E-cadherin OE MCF-7 cells. Upon live-staining of m with TMRM,
we found that the E-cadherin KO cells had significantly higher m than the WT cells at the
micropattern center without affecting those at the edge (Fig. 3-5c-f). In contrast, the E-cadherin
overexpression further significantly reduced m at the micropattern center when compared to
WT cells (Fig. 3-5c-f). Interestingly, E-cadherin overexpression resulted in an overall decrease in
m at the micropattern edge (Fig. 3-5f), which is consistent with the observation of AJ formation
80
by the E-cadherin OE cells in this region (Fig. 3-5b). These results further reinforce the essential
role of E-cadherin in negatively regulating m of cancer cell in our micropatterned tumor model.
Figure 3-5. Effects of E-cadherin knockout (KO) and overexpression (OE) on m in MCF-7
micropatterns. (a) Confocal imaging showing the localization and level of extracellular E-cadherin
(immunostained by DECMA-1) at the centers and edges of unconfined micropatterns formed by
WT and E-cadherin KO MCF-7 cells on day 4. (b) Confocal imaging showing the localization and
level of E-cadherin by GFP signal and immunostaining (24E10) at the centers and edges of
81
micropatterns formed by WT and E-cadherin OE MCF-7 cells on day 4. (c) Spatial distribution of
m indicated by TMRM fluorescence in unconfined micropatterns of WT, E-cadherin KO and
E-cadherin OE MCF-7 cells on day 4. (d) Radial distribution of m in micropatterns shown in
(c). Data points under the solid line are statistically higher in E-cadherin KO cells compared to
WT cells at each radius (p < 0.05 by 2-way ANOVA). Quantification of average TMRM
fluorescence at the centers (e) and edges (f) of micropatterns shown in (c). * P < 0.05, **P<0.01;
****P < 0.0001 in an ordinary one-way ANOVA, n = 3 micropatterns per condition.
3.4 Discussion
Loss of E-cadherin is widely known as an important step in the metastatic cascade: cancer
cells that undergo the epithelial-mesenchymal transition (EMT) lose E-cadherin, allowing them to
reduce intercellular adhesions and break off from the primary tumor
246
. On the other hand, after
tumor dissemination, the loss of E-cadherin is also associated with increased oxidative stress and
poor proliferation in the in vitro organoid tumor models
248
. However, it is unclear whether E-
cadherin loss induces oxidative stress or if the accumulation of oxidative stress potentiates the loss
of E-cadherin. Studies have shown that introduction of oxidative stress via H2O2 treatment leads
to disruption of E-cadherin mediated AJs in MCF-7 breast cancer cells and overall reduction in E-
cadherin expression in hepatocellular carcinoma cells
255,256
. On the other hand, overexpression of
E-cadherin in gastric cancer cells led to enhanced mitochondrial and glycolytic metabolism
257
.
Fragments of a different type of cadherin adhesion molecule called Fat (Ft) cadherin have been
found to directly bind to complexes of the mitochondrial electron transport chain (ETC) and
stimulate mitochondrial metabolism in Drosophila
258
. However, a mechanistic understanding of
82
whether and how E-cadherin regulates mitochondrial activity in cancer cells remains lacking. In
this study we have shown that E-cadherin expression and in particular E-cadherin mediated AJ
formation negatively regulates m in cancer cells. The present study highlights a novel pathway
wherein confinement cues from the TME regulate the m. Further studies are needed to
investigate the mechanisms and molecular adaptors by which E-cadherin expression could regulate
m, and its functional implications on cancer cell behavior.
83
Chapter 4: Regulation of Mitochondrial Membrane Potential by
YAP in Cancer Cells
4.1 Rationale
Altered mitochondrial membrane potential (ΔΨm) of cancer cells is associated with a range
of cell behaviors including decreased susceptibility to apoptosis
12
and the acquisition of an
invasive phenotype
13
. The ΔΨm is thus associated with the development of important cancer cell
hallmarks
2,14
, and a greater understanding of the mechanisms by which it alters cell phenotypes
will help in the discovery of novel therapeutic targets.
The ΔΨm is known to be regulated by the cells’ cytoskeletal elements (microtubules
85
,
microfilaments
92,96
and intermediate filaments
104,105
, and the phosphorylation state and activity of
mitochondrial proteins including uncoupling proteins
72,109
. On the other hand, the ΔΨm influences
important cell functions such as mitochondrial quality control
39
, production of reactive oxygen
species
32
and energy demand based intracellular mitochondrial transport
36
. However, the exact
mechanisms by which the ΔΨm orchestrates specific changes in cell phenotypes (such as the
acquisition of an invasive phenotype in cancer cells) remain unclear. The activation of
transcriptional co-activators Yes-Associated Protein (YAP) and Transcriptional Co-activator with
PDZ-binding Motif (TAZ) has been associated with cancer cell hallmark phenotypes such as
apoptosis evasion
259
, enhanced invasiveness and migration
260
. YAP overexpression in vitro leads
to the acquisition of a transformed phenotype with anchorage-independent growth in mammary
epithelial cells
261
. Another study reported that TAZ expression is necessary for tumorigenicity in
a mouse model of breast cancer
262
. Moreover, YAP expression has been associated with poor
prognosis in several cancer types including metastatic breast cancer
262,263
. Given the numerous
studies reporting the dysregulation of these transcription co-activators in all steps of cancer
84
progression from tumorigenesis to circulating tumor cell (CTC) survival and metastasis
260
, is it
possible that the ΔΨm exerts its effects on cancer cell phenotype by regulating the activation of
YAP? We have reported in our earlier studies that the expression of E-cadherin adhesions
negatively regulates the ΔΨm in micropatterned tumor models of MCF-7 breast cancer cells, which
was found at the micropattern centers (Ch. 3). Moreover, these regions of high-E-cadherin
adhesion and low-ΔΨm also corresponded with greater cytoplasmic localization of YAP.
Interestingly, the forced expression of E-cadherin in low E-cadherin expressing MDA-MB-231
breast cancer cells has been found to increase cytoplasmic YAP localization in these cells
264
. Thus,
in this study we aimed to investigate whether YAP could act downstream of E-cadherin adhesions
to directly regulate the ΔΨm.
We have reported in our recent studies, a strong correlation between the spatial distribution
of YAP/TAZ activity and ΔΨm in cancer cells (Ch. 2). Other studies performed with a wide range
of cell lines including myoblasts
265
, thyroid carcinoma cells
266
, gastric
267
and lung cancer cells
268
have shown that the knockdown of YAP expression leads to a decrease in the ΔΨm of these cells.
Further, in endothelial cells, it was found that palmitic acid, which inhibits YAP (by promoting its
phosphorylation and consequent nuclear exclusion), also decreases their ΔΨm
269
. Collectively,
these studies suggest that YAP activity could regulate ΔΨm. Conversely, YAP overexpression in
MDA-MB-231 breast cancer cells is reported to increase their ΔΨm
222
. These studies suggest that
YAP activity can potentially regulate cellular ΔΨm. On the other hand, studies with drugs that
inhibit ΔΨm were shown to affect the activity of YAP. Carbonyl cyanide m-chlorophenyl
hydrazine (CCCP), a well-known uncoupler of mitochondrial ETC and ATP production and
dissipator of ΔΨm
270
is reported to decrease YAP/TAZ expression in multiple myeloma cells
271
.
These results suggest the possibility of the ΔΨm regulating YAP activity. However, the exact
85
mechanisms of whether and how the ΔΨm can act as an upstream regulator of YAP activity in
cancer cells, remain unclear.
In this study, we used YAP overexpression (OE) and knockdown (k/d) models to
investigate the effects of YAP on the ΔΨm of MCF-7 cells. Using an EGFP-YAP expressing cell
line we were able to temporally resolve the effects of YAP inhibition or activation on the ΔΨm, as
well as that of ΔΨm depolarization on YAP localization. We find that YAP OE (which increases
both nuclear and cytoplasmic YAP) decreases ΔΨm. We also reveal that YAP inhibition (increase
in cytoplasmic YAP) decreases ΔΨm whereas YAP activation increases ΔΨm. Further, we also
tested the presence of YAP in mitochondrial protein fractions and showed how YAP could regulate
ΔΨm by repressing mitochondrial transcription. We thus demonstrate a novel mechanistic
understanding of how YAP could regulate the ΔΨm of cancer cells.
4.2 Materials and Methods
4.2.1 Cell Culture
MCF-7 cells were obtained from American Type Culture Collection (ATCC) and cultured
in Dulbecco’s Modified Eagle Medium (DMEM) with 10% Fetal Bovine Serum (FBS), 100 U/mL
penicillin and 100 ug/mL streptomycin.
4.2.2 Generation of EGFP-YAP Expressing and YAP Knockdown Cell Lines
Plasmid DNA encoding EGFP-C3-hYAP1 was obtained from Addgene (Addgene plasmid
# 17843 deposited by Marius Sudol ; http://n2t.net/addgene:17843 ; RRID:Addgene_17843)
272
.
Plasmid DNA was amplified with DH5α (Thermo Fisher) and isolated using the QIAprep Spin
86
Miniprep Kit (Qiagen) according to the manufacturer’s instructions, and the sequence was
confirmed by Sanger sequencing with CMV-Forward, EGFP-C-For and SV40pA-R primers at
GENEWIZ. To generate the S127A mutation in the EGFP-YAP construct, the QuikChange
Lightning Multi Site-Directed Mutagenesis Kit (Agilent Technologies) was used as per
manufacturer’s instructions, along with the following mutagenic primers: forward primer
AGCATGTTCGAGCTCATGCCTCTCCAGCTTCTC and reverse primer
GAGAAGCTGGAGAGGCATGAGCTCGAACATGCT. Mutated plasmid DNA was amplified
in XL-10 Gold Ultracompetent Cells (Agilent Technologies), isolated and screened by Sanger
sequencing as described above to confirm the S127A mutation. 1 x 10
6
MCF-7 cells were seeded
in T-75 flasks and on day 2, transfected with EGFP-YAPwt or EGFP-YAP-S127A plasmid DNA
(2 μg) using the Effectene Transfection Reagent (Qiagen). 48 hours post transfection, cells were
passaged into selective media (1 mg/ml Geneticin, Thermo Fisher). After antibiotic selection and
expansion, cells were also FACS sorted for EGFP positivity. For generating the YAP knockdown,
20,000 MCF-7 cells/well were seeded in a 24-well plate and incubated overnight. The next day
they were transduced with YAP1 MISSION shRNA Lentiviral Transduction Particles
(TRCN0000107265 and TRCN0000300282), with a multiplicity of infection (MOI) of 2 and 8
µg/ml polybrene. Cell culture media was refreshed 24 hours after the transduction, and on day 3
antibiotic selection (0.5 µg/ml puromycin; Fisher Scientific) was started. Selection was continued
till all cells in the control wells were non-viable and then extended for two cell passages. YAP
knockdown efficiency was confirmed with western blotting.
4.2.3 Assessing mitochondrial membrane potential using flow cytometry
87
100,000 MCF-7 cells/well were seeded in 6-well plates and assessed for DilC1(5)
fluorescence using the MitoProbe
TM
DilC1(5) Assay Kit for Flow Cytometry (Invitrogen)
according to manufacturer’s instruction on day 1, 3 and 5 after cell seeding. DilC1(5) stained cells
were resuspended in PBS and stained for dead cells using 1 µM of SYTOX Blue Dead Cell Stain
(Invitrogen) 5 minutes prior to analysis on the flow cytometer.
4.2.4 Live-tracking mitochondrial membrane potential and EGFP-YAP
MCF-7 cells were seeded (79,000/well) in a 12-well glass bottomed well plate (CellVis,
P12-1.5H-N) and incubated overnight. The next day, cells were rinsed with warm PBS and
incubated in 2 mL of Extracellular Imaging Buffer (130 mM sodium chloride, 5 mM potassium
chloride, 1.5 mM calcium chloride, 1 mM magnesium chloride, 25 mM 4-(2-hydroxyethyl)-1-
piperazineethanesulfonic acid (HEPES), 1 mg/mL BSA, and 5 mM glucose, with the pH adjusted
to 7.4) with 10 nM tetramethylrhodamine methyl ester (TMRM, Life Technologies) and 1 µg/mL
Hoechst 33342 (Invitrogen) for 45 min. Following this, the cells were treated with 0.5 or 5 µM
Carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (FCCP, Millipore Sigma), 5 µM
verteporfin (MedChem Express) or 10 µM N-(3-benzylthiazol-2(3H)-ylidene)- 1H-pyrrolo[2,3-
b]pyridine-3-carboxamide (TRULI, CSNpharm), and live imaged every 15 minutes for 6 hours.
4.2.5 Mitochondrial fractionation
MCF-7 cells were cultured and expanded in T-175 flasks, and mitochondrial isolation was
carried out as per manufacturer’s instructions using the Mitochondria Isolation Kit for Cultured
Cells (Thermo Scientific). The mitochondrial pellet was resuspended in 100 µL of 2% 3-((3-
88
cholamidopropyl) dimethylammonio)-1-propanesulfonate (CHAPS, Thermo Scientific) and along
with the cytoplasmic fraction stored at -80
0
C till further use for protein quantification using the
BCA assay or for western blotting.
4.2.6 Western Blotting
Protein extraction for western blotting (whole cell) was carried out from confluent T-75
flasks cultured with MCF-7, and YAP k/d MCF-7 cells (shYAP) cells. Briefly, the T-75 flasks
were placed on ice, cells rinsed with ice cold PBS and 0.5 mL of ice-cold 1.5x Laemmli buffer (2x
Laemmli sample buffer is comprised of 4% sodium dodecyl sulfate, 20% glycerol and 120 mM
Tris-Cl; recipe from Cold Spring Harbor Protocols) was added. Following this, cells were scraped
using a sterile cell scraper into a collection tube, heated at 95C for 15 minutes, centrifuged at
14,000 rpm for 5 minutes and the supernatants were aliquoted and stored at -80C till further use.
Prior to running the western blots, protein samples were quantified using the Pierce
TM
BCA Protein
Assay Kit (Thermo Scientific) according to manufacturer’s instructions. For western blotting, 20
µg of protein (whole cell lysate or cytoplasmic fraction) or 20-30 µL of mitochondrial protein was
used for polyacrylamide gel electrophoresis (PAGE) in 4-15% Mini-PROTEAN TGX
TM
Precast
Protein Gels (Bio-Rad). After blotting onto 0.2 m pore size Immun-Blot polyvinylidene
difluoride (PVDF) membranes (Bio-Rad), membranes were blocked with 5% milk in tris buffered
saline with 0.1% Tween-20 (TBST) for 1 hour at room temperature. Primary antibodies used: YAP
(sc-101199, SCBT, 1:200), TOM20 (sc-17764, SCBT, 1:100), -actin (D6A8, mAb #8457, CST,
1:1000). These were diluted at the indicated concentrations in 5% bovine serum albumin BSA
(TBST) and membranes were incubated with primary antibodies overnight at 4
0
C. Secondary
antibodies used were: anti-rabbit IgG-HRP (sc-2357, SCBT, 1:1000), m-IgG BP-HRP (sc-
89
516102, SCBT, 1:1000) and m-IgG BP-HRP (sc-516132, SCBT, 1:1000). These were diluted in
5% milk-TBST at indicated concentrations for 1 hour at room temperature. The SuperSignal™
West Pico PLUS Chemiluminescent Substrate (Thermo Scientific) was used for HRP detection.
4.2.7 Gene expression assays (qPCR)
RNA was isolated from MCF-7 and MCF-7 shYAP (YAP k/d) cells using the PicoPure™
RNA Isolation Kit (Applied Biosystems) according to manufacturer’s instructions. 1.5 g of RNA
of each cell line was used for cDNA synthesis using the SuperScript IV VILO Master Mix
(Invitrogen). TaqMan™ Universal Master Mix II, with UNG (Applied Biosystems) was used to
prepare the qPCR reactions along with the TaqMan probes listed in Table 4-1.
4.3 Results
4.3.1 YAP overexpression decreases the m of MCF-7 breast cancer cells
We have previously reported a spatial distribution of m in micropatterns of MCF-7
breast cancer cells where cells at the edges of the micropatterned tumor island have a 3-fold higher
m compared to cells at the micropatterned center (Ch. 2, Ch. 3). While investigating the
mechanisms by which MCF-7 cell m was differentially regulated in the micropatterns, we found
that the formation of cadherin mediated adherens junctions (AJs) at the micropattern centers
negatively regulated m (Ch. 3). The activation of signaling pathways downstream of E-cadherin
mediated AJ formation has been reported to regulate the localization (and hence activity) of YAP
264,273
. Kim et al., have shown that the expression of E-cadherin (but not mutant E-cadherin that is
90
incapable of binding to endogenous - and -catenin) in MDA-MB-231 breast cancer cells leads
to cytoplasmic localization of YAP at high cell densities, which is not seen without E-cadherin
expression
264
. Moreover, in our own studies, we reported an association of low- m at the
micropattern centers with cytoplasmic YAP localization that of high- m at the micropattern
edges with nuclear YAP localization (Fig. 2-4). These results led us to hypothesize that YAP
activity could directly regulate the m of cancer cells.
To investigate the effects of YAP on the m of cancer cells, we first overexpressed two
forms of EGFP-YAP in MCF-7 cells: regular YAP1 (referred to as ‘YAPwt’) and a mutated form
of YAP1 that prevents its phosphorylation at Serine 127 (referred to as ‘S127A’). YAP
phosphorylation at Ser-127 is a result of the activity of the Hippo pathway effector Lats-kinase
and promotes its cytoplasmic localization
274,275
. We then stained un-patterned non-transfected
MCF-7 cells along with the YAP OE cells with DilC1(5), a m dye
276
and analyzed fluorescence
using flow cytometry. We confirmed that DilC1(5) fluorescence observed was abrogated with the
treatment of CCCP (a known m depolarizer
138
) indicating that it is representative of m (Fig.
4-1a). Next, we see that the m of YAP OE cells (YAPwt and S127A) is significantly lower than
that of non-transfected MCF-7 cells (Fig. 4-1b-d). Interestingly, we see that in the overall YAP
expression (as measured by immunostaining with YAP antibody), there is a significantly greater
than 2-fold increase in the nuclear as well as cytoplasmic YAP fluorescence in both the YAP OE
cells when compared to the MCF-7 control (Fig. 4-1e-f). Since we have earlier reported a positive
correlation of nuclear YAP with m (Fig. 2-4d), these results suggest that the increase in
cytoplasmic YAP in the YAP OE cells could lead to the observed decrease in their m. Moreover,
we also see that the nuclear to cytoplasmic YAP ratio remains conserved with YAP OE (except
for a slight decrease in this ratio for the S127A cells) (Fig. 4-1g).
91
Figure 4-1: Effect of YAP overexpression (OE) on the m of MCF-7 cells. (a) Average DilC1(5)
fluorescence of MCF-7 control and YAP OE cells with and without CCCP treatment
92
(representative dataset). (b) Average DilC1(5) fluorescence of MCF-7 control and YAP OE cells
from a representative dataset. (c) Quantification of the results shown in (b), N = 10 independent
experiments, ** P < 0.0021 in a RM one-way ANOVA. (d) Pair-wise comparisons of the DilC1(5)
fluorescence across the different cell types. ** P < 0.0021 in a paired t test. (e)
Immunofluorescence of sparsely seeded MCF-7 control, and YAP OE cells stained with a YAP
antibody (red), along with the EGFP-YAP signal (green) expressed by the YAP OE cells and
nuclear staining (DAPI, blue). (f) Quantification of nuclear and cytoplasmic YAP fluorescence in
cells shown in (e) using ImageJ. (g) Quantification of nuclear to cytoplasmic ratio of YAP
fluorescence intensity in cells shown in (e). f-g: * P < 0.0332, *** P < 0.0002 and **** P < 0.0001
in an Ordinary one-way ANOVA.
4.3.2 YAP inhibition by verteporfin decreases m whereas YAP activation by TRULI
increases the m of MCF-7 cells
We next aimed to investigate the effect of YAP inhibition and activation on the m of
MCF-7 cells. To visualize the effect of YAP inhibition and activation on YAP localization in real-
time, we used live imaging to track EGFP-YAP in real-time upon the addition of YAP inhibitory
and activation stimuli using the YAP OE cells as described earlier. To inhibit YAP activity, we
treated MCF-7 cells with 5M verteporfin. Verteporfin was found to inhibit the interaction of YAP
with TEAD, thereby reducing nuclear YAP activity
277
. Another study reported that verteporfin
promotes cytoplasmic localization of YAP by increasing the expression of 14-3-3 protein
278
. We
observed a rapid depolarization of the m in both the control as well as the YAP OE cells as soon
as within 15 minutes of treatment with verteporfin (Fig. 4-2a-b). We also observed that verteporfin
treatment led to an increase in the cytoplasmic YAP fluorescence (EGFP-YAP) up to 30 minutes,
93
followed by a decline in the fluorescence signal (Fig. 4-2a, c). This is consistent with reports of
verteporfin increasing the sequestration of cytoplasmic YAP along with an increase in 14-3-3
protein levels that eventually leads to its degradation
278
. It is interesting to note that the m started
to decrease along with the increase in cytoplasmic YAP fluorescence up to 30 mins, after which it
continued to remain low even after the EGFP-YAP signal faded. This suggests that an increase in
cytoplasmic YAP or the overall YAP inhibition could down-regulate the m of cancer cells. On
the other hand, we tested the effects of activating YAP on the m by treatment with N-(3-
benzylthiazol-2(3H)-ylidene)-1H-pyrrolo[2,3-b]pyridine-3- carboxamide (TRULI) which was
recently discovered as a non-toxic small molecule inhibitor of Lats-kinase that prevents YAP
phosphorylation
279
. We see that YAP activation (by inhibiting YAP phosphorylation) using
TRULI led to an increase in the overall EGFP-YAP fluorescence which was also associated with
a sustained and significant increase in the m (Fig. 4-2a-c). Overall, these results strongly
suggest that YAP activation up-regulates m, whereas YAP inhibition down-regulates the m
of cancer cells.
94
Figure 4-2: Effect of YAP inhibition (verteporfin) and activation (TRULI) on the m of MCF-7
95
cells. (a) Time-lapse microscopy showing representative images for TMRM fluorescence (red;
m) and EGFP-YAP (green) at 0, 30 and 60 minutes after the addition of 5 M verteporfin (VP),
10M TRULI or control for MCF-7 non-transfected control cells as well as the YAP OE cells
(YAPwt and S127A). (b) Average TMRM fluorescence intensity over time for the conditions
shown in (a). Average cytoplasmic EGFP-YAP intensity (c) and nuclear EGFP-YAP intensity (d)
in the YAP OE cells.
4.3.3 Depolarization of the m does not alter EGFP-YAP fluorescence intensity and
localization
Our results so far suggested that YAP activity regulates the m of MCF-7 cells. Next, we
aimed to investigate whether changes in the m could regulate YAP activity. To test the effect
of m depolarization on YAP, we live imaged EGFP-YAP in MCF-7 YAP OE cells upon
treatment with 0.5 µM and 5 µM FCCP (which is a known depolarizer of the m
138
). Treatment
with 5 µM FCCP led to a decline of m within 2 hours of treatment, whereas treatment with 0.5
µM FCCP did not depolarize the m completely (Fig. 4-3a, b). Interestingly, we found that even
after a long period of sustained m depolarization with 5 µM FCCP treatment, there were no
significant changes in cytoplasmic and nuclear EGFP-YAP fluorescence intensities compared to
the untreated controls in the MCF-7 YAP OE cells (Fig. 4-3a, c-d). This strongly suggests that
m depolarization does not regulate YAP activity in MCF-7 cells (at least over a period of 6
hours), however, any changes in YAP activity (increase in cytoplasmic YAP by VP or increase in
nuclear YAP by TRULI) immediately result in m changes.
Additionally, we also observe that the FCCP treatment led to a dose dependent increase in
the baseline FITC autofluorescence in MCF-7 non-transfected control cells (as shown in Fig. 4-
96
S1 b), which was comparable to the increase in FITC autofluorescence upon treatment with
TRULI, but very different from its steep increase upon verteporfin treatment induced m
depolarization (Fig. 4-S1 a). FITC autofluorescence is reported to be a marker of senescence in
vitro
280
, suggesting that inhibiting YAP through verteporfin treatment leads to a very distinct type
of mitochondrial inhibition that initiates cell senescence.
97
98
Figure 4-3: Effect of FCCP-induced m depolarization on cytoplasmic and nuclear YAP
fluorescence. (a) Time-lapse microscopy showing representative images for TMRM fluorescence
(red; m) and EGFP-YAP (green) at 0, 3 and 6 hours after the addition of 0.5 M and 5 M
FCCP (or untreated control) for MCF-7 non-transfected control cells as well as the YAP OE cells
(YAPwt and S127A). (b) Average TMRM fluorescence intensity over time for the conditions
shown in (a). Average cytoplasmic EGFP-YAP intensity (c) and nuclear EGFP-YAP intensity (d)
in the YAP OE cells upon FCCP treatment over a 6-hour time period.
4.3.4 YAP can be detected in the mitochondrial protein isolates of MCF-7 cells
Next, we aimed to investigate how YAP regulates the m and hypothesized that YAP
could localize to the mitochondria in order to regulate the m. We isolated mitochondrial protein
from MCF-7 cells and used western blotting to test for the presence of YAP in them. As shown in
Fig. 4-4, we were able to visualize a clear YAP band in the mitochondrial protein fraction. Probing
with the -actin antibody revealed that there was strong presence of -actin within the
mitochondrial protein fraction, thus raising concerns of cytosolic protein contamination in the
isolated mitochondrial protein (Fig. 4-4a). However, since actin is known to be closely associated
with the mitochondrial outer surface, we repeated the western blot, using TFAM as a mitochondrial
marker (which was present in the mitochondrial protein and absent in the cytosolic protein) and
GAPDH as a cytosolic marker (Fig. 4-4b). We saw minimal presence of GAPDH in the
mitochondrial fraction compared to the cytosolic fraction indicating that the mitochondrial fraction
was mostly free from cytosolic proteins (Fig. 4-4b). To confirm these results, we performed
another western blot with mitochondrial and cytosolic protein fractions from both MCF-7 and
YAP k/d MCF-7 (shYAP) cells (Fig. 4-4c), keeping the amount of protein loaded constant (20
99
g). We observed that there was a decrease in YAP in both the mitochondrial and cytoplasmic
protein fractions in the YAP k/d cells. These results suggest that a potential mechanism by which
YAP could regulate the m of MCF-7 cells is by localizing to the mitochondria. Whether the
mitochondrially localized YAP is phosphorylated (like the cytoplasmic YAP) remains to be
investigated. We next performed a proximity ligation assay (PLA) using YAP OE (YAPwt) MCF-
7 cells to determine whether YAP interacted with mitochondrial transcription factors (we tested
the interaction of YAP with mitochondrial transcription factor A; TFAM). The formation of PLA
puncta in this assay (shown in red, Fig. 4-4d) is representative of close association and potential
protein-protein interaction of the targets tested (YAP and TFAM). Interestingly, we observed a
significant increase the number of PLA puncta relative to the single antibody control condition
(representing non-specific signal). These results suggest that the mitochondrially localized YAP
could interact with TFAM to regulate mitochondrial transcription. However, we recognize that the
PLA test has its limitations due to the high background non-specific signal as seen with the single
antibody control and these results will have to be validated using other experimental approaches
such co-immunoprecipitation (co-IP) that are more traditionally used to detect and confirm
protein-protein interactions.
100
Figure 4-4: Identification of YAP in mitochondrial protein of MCF-7 cells. Western blot with
mitochondrial and cytoplasmic and protein fractions isolated from MCF-7 cells stained with YAP
(63.7) and -actin (a) or YAP (D8H1X), TFAM and GAPDH (b) in MCF-7 cells. (c) Western blot
with cytosolic and mitochondrial protein fractions isolated from MCF-7 and YAP k/d MCF-7 cells
(shYAP), stained with YAP (D8H1X), TFAM and GAPDH. (d) Maximum intensity projections
of z-stack confocal images after a Proximity Ligation Assay (PLA) showing nuclei (blue) and PLA
puncta (red) in MCF-7 YAPwt cells with both YAP and TFAM antibodies or the single antibody
(YAP) control (non-specific signal), along with a quantification of the PLA puncta. * P < 0.0332
in an unpaired t-test.
101
4.3.5 YAP k/d leads to an increase in mitochondrial transcription of ETC genes
Since YAP was detected within the mitochondria, we next wanted to investigate the effects
of YAP on mitochondrial transcription and nuclear transcription of mitochondrial electron
transport chain (ETC) genes. First, we knocked down YAP in MCF-7 cells using an shRNA
approach (see methods section) and validated the knockdown by western blot (~50% knockdown
efficiency, Fig. 4-5a). Next, we isolated RNA from MCF-7 control and YAP k/d (‘shYAP’) cells
and performed a qPCR assay. We selected a panel of mitochondrially encoded genes as described
by Roedder el al.
281
. Among the mitochondrially transcribed genes we tested, we found a
significant up-regulation of genes that encoded for subunits of ETC complexes IV and V as well
as a gene that encodes for mitochondrial ribosomal RNA in the YAP k/d MCF-7 cells (Fig. 4-5c).
This suggests that the presence of YAP in the mitochondria could be negatively regulating
mitochondrial transcription. Interestingly, we also found that there was also a significant up-
regulation of nuclear transcribed genes that encode for subunits of mitochondrial ETC complexes
I-II (Fig. 4-5d). On the other hand, there was no significant difference in gene expression of other
nuclear encoded genes regulating mitochondrial activity including those that regulate
mitochondrial fusion, fission and biogenesis (PGC1A, PGC1B, Fig. 4-5e). Interestingly however,
we did find a significant upregulation in NRF1 which also regulates mitochondrial biogenesis
282
in the YAP k/d cells. Further, we also observed no significant differential regulation in the gene
expression of glycolytic genes except for PDHA1 which encodes for a subunit of pyruvate
dehydrogenase (Fig. 4-5f). Pyruvate dehydrogenase converts pyruvate (the product of glycolysis)
to acetyl co-A (substrate for the mitochondrial TCA cycle) and thus regulates the entry of
metabolites from glycolysis to oxidative metabolism
283
. This data suggests that the presence of
YAP downregulates mitochondrial metabolism in MCF-7 cells. Moreover, reducing mitochondrial
102
transcription might be a potential mechanism by which the mitochondrially localized YAP
regulates the m of MCF-7 cells.
103
Figure 4-5: Effect of YAP k/d on mitochondrial transcription. (a) Western blot showing YAP and
104
β-actin staining of cell lysates from MCF-7 (control) and YAP k/d (shYAP) cells. (b) Relative
YAP expression in the MCF-7 YAP k/d cells (shYAP) normalized to β-actin. qPCR assay showing
log2fold change in the gene expression of MCF-7 YAP k/d cells relative to MCF-7 control, for
mitochondria encoded genes (c), nuclear encoded mitochondrial ETC genes (d), nuclear encoded
genes regulating mitochondrial activity (e) and nuclear encoded genes regulating glycolysis (f).
c-d: N = 3 independent experiments, e-f: N = 2 independent experiments. ** P < 0.0021 and * P
< 0.0332 in a one sample t test.
4.3.6 YAP overexpression decreases mitochondrial respiration and increases glycolysis in
MCF-7 cells
The gene expression data suggested that YAP downregulates the transcription of
mitochondrial ETC genes since the knockdown of YAP significantly increased mitochondrial
transcription. Next, we wanted to test the functional consequence of YAP expression on
mitochondrial respiration. To investigate this, we used the YAP OE cells, and performed a
mitochondrial stress test on a Seahorse analyzer to record mitochondrial respiration (Fig. 4-6a). In
a mitochondrial stress test, the oxygen consumption rate (OCR) of cells seeded in a specialized
Seahorse assay plate is first measured without any drug treatment to get an estimate of the basal
respiration of the cells (Fig. 4-6a). Next, treatment with oligomycin inhibits the ATP synthase
(ETC Complex V) which leads to a decrease in OCR, and the resultant OCR is a measure of the
oxygen consumed by the cells due to a proton leak across the inner mitochondrial membrane
(IMM). The cells are then treated with FCCP which permeabilizes the IMM to H
+
ions, leading to
maximal oxygen consumption by the cells, following which rotenone and anytimycin treatment
shuts down the ETC and all mitochondrial oxygen consumption. In this manner, the mitochondrial
105
stress test measures several important mitochondrial functional parameters. We found that the
overall OCR of MCF-7 cells overexpressing YAP (YAPwt and S127A) was lower than that of the
control MCF-7 cells (Fig. 4-6b). Interestingly, both non-mitochondrial respiration (which takes up
a small fraction of the total respiration) as well as mitochondrial respiration were significantly
decreased in the YAP OE cells (Fig. 4-6c, d-f). Additionally, the decrease in basal respiration of
YAP OE cells resulted from decreased proton leak and ATP production when compared to control
MCF-7 cells (Fig. 4-6d-e, g). There is however no significant difference in the coupling efficiency
in the MCF-7 cells upon YAP OE, indicating that the YAP OE does not uncouple oxygen
consumption from ATP production (Fig. 4-6h). It is interesting to note as per previous reports that
the spare respiratory capacity of the MCF-7 cells is very low
284
and decreases with the OE of YAP
(YAPwt) (Fig. 4-6i). Together, this data shows how YAP overexpression decreases mitochondrial
respiration in MCF-7 cells, which is in agreement with the gene expression studies that show an
increase in mitochondrial ETC transcription upon YAP k/d. This suggests that increased
cytoplasmic YAP may play a role in down regulating the m of cancer cells.
106
Figure 4-6: Effect of YAP OE on mitochondrial bioenergetics of MCF-7 cells. (a) Schematic
showing the addition of different mitochondrial inhibitors during a mitochondrial stress test using
the Seahorse XF analyzer. (b) Normalized oxygen consumption rate of MCF-7 cells
overexpressing YAP compared to non-transfected control cells. Non-mitochondrial respiration (c),
basal respiration (d), ATP production (e), maximal respiration (f), proton leak (g), coupling
107
efficiency (h) and spare respiratory capacity (i) of MCF-7 YAP OE cells compared to control, as
calculated by the mitochondrial stress test. * P < 0.0332, ** P < 0.0021, *** P < 0.0002 and ****
P < 0.0001 in a one-way ANOVA.
4.4 Discussion
YAP is a well-known transcriptional co-activator that regulates several cell functions
including cell proliferation and differentiation
285
, metabolic pathways such as glycolysis and lipid
metabolism
286
. Different types of mechanical stimuli (substrate stiffness, cell-ECM contact area,
cell crowding)
287
in addition to metabolic stimuli such as glycolytic enzymes regulate YAP
activity
288
. However, mechanisms by which YAP regulates mitochondrial activity and metabolism
remain unclear. Since we have previously described how mechanical cues from the TME could
regulate the m of cancer cells, and that YAP subcellular localization correlates with spatial
gradients of m, we hypothesized that YAP could regulate mitochondrial activity downstream
of mechanical stimuli received by the cancer cells.
There are reports of YAP k/d reducing the m of several cell types
265-268
, suggesting that
YAP positively regulates the m. In a YAP OE model in MDA-MB-231 breast cancer cells, it
was reported that the YAP OE increases mitochondrial mass and mitochondrial ETC complex
V
222
. However, when we overexpressed YAP in MCF-7 cells, we found that it significantly
decreased the m (Fig. 4-1). Overexpression of YAP in MCF-7 cells led to an increase in both
nuclear as well as cytoplasmic YAP when compared to the control cells. It is thus possible that
nuclear and cytoplasmic YAP differentially regulate the m. Indeed, we did identify YAP in the
mitochondrial protein fraction of MCF-7 cells (Fig. 4-4), suggesting that there is mitochondrial
localization of YAP. When we tested mitochondrial gene expression of a YAP k/d, we found that
108
there was a significant increase in the expression of mitochondrial ETC complex IV, V subunits
as well as mitochondrial rRNA (Fig 4-5). This data suggests that the presence of YAP in the
mitochondria could be inhibiting mitochondrial transcription. Kim et al. have reported that YAP
could also serve as a transcription co-repressor and have identified around 100 gene targets whose
transcription is inhibited by (nuclear) YAP
289
. Together with our data, this suggests that
mitochondrially-localized YAP could function as a repressor of mitochondrial transcription.
Further studies are needed to confirm the identity of mitochondrial proteins that YAP interacts
with. Interestingly, Liu et al. have reported ATP synthase subunits ( and ) and Tu translational
elongation factor, mitochondrial (TUFM) to be among the proteins that co-interact with YAP and
another transcription factor (TFCP2)
290
. These studies show the possibility of direct YAP
interactions with mitochondrial proteins, and it would also be interesting to investigate which
forms of YAP (Hippo pathway mediator phosphorylated or non-phosphorylated) localize to the
mitochondria. Since the m is the most downregulated in conditions where there is cytoplasmic
YAP (at micropattern centers), it is likely that some fraction of the phosphorylated and thus
cytoplasmic YAP localizes to the mitochondria to repress mitochondrial activity.
Further supporting this conclusion, we find that treatment with verteporfin, which leads to
an increase in the cytoplasmic YAP (Fig. 4-2), leads to a decrease in the m whereas treatment
with TRULI which is an activator of nuclear YAP (works by preventing YAP phosphorylation by
Lats) leads to an increase in the m (Fig. 4-2). It is possible that the m increase upon TRULI
treatment is due to a decrease in the YAP phosphorylation and thus a decrease in phospho-YAP
mediated mitochondrial inhibition. We also see that modulation of YAP activity (either using
verteporfin or TRULI) leads to immediate changes in the m (within 15-30 minutes, Fig. 4-2)
whereas changes in m (depolarization by FCCP) have no significant impact on YAP localization
109
for up to 6 hours (Fig. 4-3). Fan et al. report that prolonged 5 M CCCP treatment (48 hours)
reduces YAP expression in multiple myeloma cells. Together with our data, this suggests that
while YAP regulates m in a relatively short time frame (30 mins), the regulation of YAP by
changes in mitochondrial activity could take much longer (48 hours).
The negative regulation of m by cytoplasmic YAP is also supported by the
mitochondrial stress test data which shows that YAP OE leads to a decrease in mitochondrial
respiration of MCF-7 cells (Fig. 4-6). Not surprisingly, these cells had a compensatory increase in
glycolysis as measured by the extracellular acidification rate (ECAR) in a glycolysis stress test
(Fig. 4S-1). These results also complement a recent study by White et al., who show that YAP k/d
leads to decreased glycolysis (glycolysis stress test) and increased mitochondrial respiration
(mitochondrial stress test) in renal carcinoma cells
291
. The data presented here thus provides
compelling evidence for negative regulation of m by mitochondrial YAP localization in cancer
cells. Future studies exploring the mechanisms of YAP mitochondrial localization will help shed
light on novel pathways by which YAP could regulate mitochondrial metabolism.
110
Figure 4S-1: Effect of YAP OE on glycolytic metabolism of MCF-7 cells. (a) Schematic showing
the addition of different glycolysis substrates and inhibitors during a glycolysis stress test using
the Seahorse XF analyzer. (b) Normalized extracellular acidification rate (ECAR) of MCF-7 cells
overexpressing YAP compared to non-transfected control cells. Non-glycolytic acidification (c),
glycolysis (d), glycolytic capacity (e) and glycolytic reserve (f) of MCF-7 YAP OE cells compared
to control, as calculated by the glycolysis stress test. * P < 0.0332, ** P < 0.0021, *** P < 0.0002
and **** P < 0.0001 in a one-way ANOVA.
111
Figure 4S-2: Effect of drug treatments on FITC autofluorescence of MCF-7 cells. (a) FITC
autofluorescence representative images (5 M verteporfin) and quantification in MCF-7 cells
treated with 5 M verteporfin (VP) or 10 M TRULI compared to control. (b) FITC
autofluorescence representative images (5 M FCCP) and quantification in MCF-7 cells treated
with 0.5 M or 5 M FCCP compared to control.
112
TaqMan Assay ID Gene Target
Hs01060665_g1 ACTB
Hs00204417_m1 NDUFA8
Hs00268117_m1 SDHB
Hs01890823_s1 UQCRB
Hs01053235_g1 COX17
Hs01081389_g1 ATP5J
Hs00606086_m1 HK2
Hs01378790_g1 LDHA
Hs01049345_g1 PDHA1
Hs00173304_m1 PPARGC1A
Hs00993805_m1 PPARGC1B
Hs00602161_m1 NRF1
Hs00966851_m1 MFN1
Hs00208382_m1 MFN2
Hs01047018_m1 OPA1
Hs01552605_m1 DNM1L
Hs00211420_m1 FIS1
Hs01066656_m1 MFF
Hs00892681_m1 SLC2A1
Hs01006127_m1 SLC16A4
Hs00943178_g1 PGK1
Hs01921501_s1 PPP1R3C
113
Hs02596861_s1 MT-7S
Hs02596865_g1 MT-CO2
Hs02596862_g1 MT-ATP6
Hs02596879_g1 MT-ND6
Hs02596878_g1 MT-ND5
Hs02596859_g1 MT-RNR1
Hs02596867_s1 MT-CYB
Hs02596874_g1 MT-ND2
Hs02596864_g1 MT-CO1
Hs02596873_s1 MT-ND1
Hs02596863_g1 MT-ATP8
Table 4-1: List of TaqMan assays used for qPCR.
114
References
1 Koppenol, W. H., Bounds, P. L. & Dang, C. V. Otto Warburg's contributions to current
concepts of cancer metabolism. Nat Rev Cancer 11, 325-337, doi:10.1038/nrc3038
(2011).
2 Hanahan, D. & Weinberg, R. A. Hallmarks of cancer: the next generation. Cell 144, 646-
674, doi:10.1016/j.cell.2011.02.013 (2011).
3 Zhu, A., Lee, D. & Shim, H. Metabolic positron emission tomography imaging in cancer
detection and therapy response. Semin Oncol 38, 55-69,
doi:10.1053/j.seminoncol.2010.11.012 (2011).
4 Ward, P. S. & Thompson, C. B. Metabolic reprogramming: a cancer hallmark even
warburg did not anticipate. Cancer Cell 21, 297-308, doi:10.1016/j.ccr.2012.02.014
(2012).
5 Ashton, T. M., McKenna, W. G., Kunz-Schughart, L. A. & Higgins, G. S. Oxidative
Phosphorylation as an Emerging Target in Cancer Therapy. Clin Cancer Res 24, 2482-
2490, doi:10.1158/1078-0432.CCR-17-3070 (2018).
6 Whitaker-Menezes, D. et al. Hyperactivation of oxidative mitochondrial metabolism in
epithelial cancer cells in situ: visualizing the therapeutic effects of metformin in tumor
tissue. Cell Cycle 10, 4047-4064, doi:10.4161/cc.10.23.18151 (2011).
7 Porporato, P. E. et al. A mitochondrial switch promotes tumor metastasis. Cell Rep 8,
754-766, doi:10.1016/j.celrep.2014.06.043 (2014).
8 Simoes, R. V. et al. Metabolic plasticity of metastatic breast cancer cells: adaptation to
changes in the microenvironment. Neoplasia 17, 671-684, doi:10.1016/j.neo.2015.08.005
(2015).
9 Dupuy, F. et al. PDK1-Dependent Metabolic Reprogramming Dictates Metastatic
Potential in Breast Cancer. Cell Metab 22, 577-589, doi:10.1016/j.cmet.2015.08.007
(2015).
115
10 Summerhayes, I. C. et al. Unusual retention. of rhodamine 123 by mitochondria in
muscle and carcinoma cells. Proc Natl Acad Sci U S A 79, 5292-5296, (1982).
11 Chen, L. B. Mitochondrial Membrane Potential in Living Cells. Annual Review of Cell
Biology 4, 155-181, doi:10.1146/annurev.cb.04.110188.001103 (1988).
12 Heerdt, B. G., Houston, M. A., Wilson, A. J. & Augenlicht, L. H. The intrinsic
mitochondrial membrane potential (δψm) is associated with steady-state mitochondrial
activity and the extent to which colonic epithelial cells undergo butyrate-mediated growth
arrest and apoptosis. Cancer Research 63, 6311-6319 (2003).
13 Heerdt, B. G., Houston, M. A. & Augenlicht, L. H. The intrinsic mitochondrial
membrane potential of colonic carcinoma cells is linked to the probability of tumor
progression. Cancer Res 65, 9861-9867, doi:10.1158/0008-5472.CAN-05-2444 (2005).
14 Hanahan, D. & Weinberg, R. A. The Hallmarks of Cancer. Cell 100, 57-70 (2000).
15 Beadnell, T. C., Scheid, A. D., Vivian, C. J. & Welch, D. R. Roles of the mitochondrial
genetics in cancer metastasis: not to be ignored any longer. Cancer Metastasis Rev 37,
615-632, doi:10.1007/s10555-018-9772-7 (2018).
16 Levine, A. J. & Puzio-Kuter, A. M. The Control of the Metabolic Switch in Cancers by
Oncogenes and Tumor Suppressor Genes. Science 330 (2010).
17 Nagarajan, A., Malvi, P. & Wajapeyee, N. Oncogene-directed alterations in cancer cell
metabolism. Trends Cancer 2, 365-377, doi:10.1016/j.trecan.2016.06.002 (2016).
18 Pavlova, N. N. & Thompson, C. B. The Emerging Hallmarks of Cancer Metabolism. Cell
Metab 23, 27-47, doi:10.1016/j.cmet.2015.12.006 (2016).
19 Alberts, B., Johnson, A. & Lewis, J. (2002).
20 Kühlbrandt, W. Structure and function of mitochondrial membrane protein complexes.
BMC Biology 13, 1-11, doi:10.1186/s12915-015-0201-x (2015).
116
21 Pelley, J. W. Citric Acid Cycle, Electron Transport Chain, and Oxidative
Phosphorylation. Elsevier's Integrated Review Biochemistry, 57-65, doi:10.1016/b978-0-
323-07446-9.00007-6 (2012).
22 Heerdt, B. G., Houston, M. A. & Augenlicht, L. H. Growth properties of colonic tumor
cells are a function of the intrinsic mitochondrial membrane potential. Cancer Res 66,
1591-1596, doi:10.1158/0008-5472.CAN-05-2717 (2006).
23 Houston, M. A., Augenlicht, L. H. & Heerdt, B. G. Stable differences in intrinsic
mitochondrial membrane potential of tumor cell subpopulations reflect phenotypic
heterogeneity. Int J Cell Biol 2011, 978583, doi:10.1155/2011/978583 (2011).
24 Bonnet, S. et al. A mitochondria-K+ channel axis is suppressed in cancer and its
normalization promotes apoptosis and inhibits cancer growth. Cancer Cell 11, 37-51,
doi:10.1016/j.ccr.2006.10.020 (2007).
25 Momcilovic, M. et al. In vivo imaging of mitochondrial membrane potential in non-
small-cell lung cancer. Nature 575, 380-384, doi:10.1038/s41586-019-1715-0 (2019).
26 Vyas, S., Zaganjor, E. & Haigis, M. C. Mitochondria and Cancer. Cell 166, 555-566,
doi:10.1016/j.cell.2016.07.002 (2016).
27 Belous, A. et al. Reversed activity of mitochondrial adenine nucleotide translocator in
ischemia-reperfusion. Transplantation 75, 1717-1723,
doi:10.1097/01.TP.0000063829.35871.CE (2003).
28 Zorova, L. D. et al. Functional Significance of the Mitochondrial Membrane Potential.
Biochemistry (Moscow) Supplement Series A: Membrane and Cell Biology 12, 20-26,
doi:10.1134/S1990747818010129 (2018).
29 Zorova, L. D. et al. Mitochondrial membrane potential. Anal Biochem 552, 50-59,
doi:10.1016/j.ab.2017.07.009 (2018).
30 Li, X. et al. Targeting mitochondrial reactive oxygen species as novel therapy for
inflammatory diseases and cancers. Journal of Hematology and Oncology 6, 1-19,
doi:10.1186/1756-8722-6-19 (2013).
117
31 Thannickal, V. J. & Fanburg, B. L. Reactive oxygen species in cell signaling. American
Journal of Physiology-Lung Cellular and Molecular Physiology 279, L1005-L1028,
doi:10.1002/9780470988565.ch9 (2000).
32 Korshunov, S. S., Skulachev, V. P. & Starkov, A. A. High protonic potential actuates a
mechanism of production of reactive oxygen species in mitochondria. FEBS Letters 416,
15-18, doi:10.1016/S0014-5793(97)01159-9 (1997).
33 Lee, I., Bender, E., Arnold, S. & Kadenbach, B. New control of mitochondrial membrane
potential and ROS formation - A hypothesis. Biological Chemistry 382, 1629-1636,
doi:10.1515/BC.2001.198 (2001).
34 Suski, J. M. et al. Vol. 810 103-117 (2012).
35 Lebiedzinska, M. et al. Oxidative stress-dependent p66Shc phosphorylation in skin
fibroblasts of children with mitochondrial disorders. Biochimica et Biophysica Acta -
Bioenergetics 1797, 952-960, doi:10.1016/j.bbabio.2010.03.005 (2010).
36 Hyde, B. B., Twig, G. & Shirihai, O. S. Organellar vs cellular control of mitochondrial
dynamics. Seminars in Cell and Developmental Biology 21, 575-581,
doi:10.1016/j.semcdb.2010.01.003 (2010).
37 Twig, G., Hyde, B. & Shirihai, O. S. Mitochondrial fusion, fission and autophagy as a
quality control axis: The bioenergetic view. Biochimica et Biophysica Acta -
Bioenergetics 1777, 1092-1097, doi:10.1016/j.bbabio.2008.05.001 (2008).
38 Chen, H. & Chan, D. C. Vol. 59 119-144 (2004).
39 Twig, G. et al. Fission and selective fusion govern mitochondrial segregation and
elimination by autophagy. EMBO Journal 27, 433-446, doi:10.1038/sj.emboj.7601963
(2008).
40 van der Bliek, A. M., Shen, Q. & Kawajiri, S. Mechanisms of mitochondrial fission and
fusion. Cold Spring Harbor Perspectives in Biology 5, doi:10.1101/cshperspect.a011072
(2013).
118
41 Meeusen, S., McCaffery, J. M. & Nunnari, J. Mitochondrial fusion intermediates
revealed in vitro. Science 305, 1747-1752, doi:10.1126/science.1100612 (2004).
42 Legros, F. d. r., Lombe`s, A., Frachon, P. & Rojo, M. Mitochondrial Fusion in Human
Cells is Efficient, Requires the Inner Membrane Potential, and Is Mediated by
Mitofusins. Molecular Biology of the Cell 13, 4343-4354, doi:10.1091/mbc.E02 (2002).
43 Liu, X., Weaver, D., Shirihai, O. & Hajnóczky, G. Mitochondrial kiss-and-run: Interplay
between mitochondrial motility and fusion-fission dynamics. EMBO Journal 28, 3074-
3089, doi:10.1038/emboj.2009.255 (2009).
44 Duvezin-Caubet, S. et al. Proteolytic processing of OPA1 links mitochondrial
dysfunction to alterations in mitochondrial morphology. Journal of Biological Chemistry
281, 37972-37979, doi:10.1074/jbc.M606059200 (2006).
45 Shirihai, O. S., Song, M. & Dorn, G. W. How mitochondrial dynamism orchestrates
mitophagy. Circulation Research 116, 1835-1849,
doi:10.1161/CIRCRESAHA.116.306374 (2015).
46 Jheng, H. F. et al. Mitochondrial Fission Contributes to Mitochondrial Dysfunction and
Insulin Resistance in Skeletal Muscle. Molecular and Cellular Biology 32, 309-319,
doi:10.1128/mcb.05603-11 (2012).
47 Zhao, J. et al. Mitochondrial dynamics regulates migration and invasion of breast cancer
cells. Oncogene 32, 4814-4824, doi:10.1038/onc.2012.494 (2013).
48 Sun, X. et al. Mitochondrial fission promotes cell migration by Ca 2+
/CaMKII/ERK/FAK pathway in hepatocellular carcinoma. Liver International 38, 1263-
1272, doi:10.1111/liv.13660 (2018).
49 Marchi, S. et al. Mitochondria-Ros Crosstalk in the Control of Cell Death and Aging.
Journal of Signal Transduction 2012, 1-17, doi:10.1155/2012/329635 (2012).
50 Mijaljica, D., Prescott, M. & Devenish, R. J. Mitophagy: An Overview. Fourth Edi edn,
Vol. 4 (Elsevier Inc., 2014).
119
51 Twig, G. & Shirihai, O. S. The interplay between mitochondrial dynamics and
mitophagy. Antioxidants and Redox Signaling 14, 1939-1951, doi:10.1089/ars.2010.3779
(2011).
52 Vives-Bauza, C., De Vries, R. L. A., Tocilescu, M. & Przedborski, S. PINK1/Parkin
direct mitochondria to autophagy. Autophagy 6, 315-316, doi:10.4161/auto.6.2.11199
(2010).
53 Tanaka, A. et al. Proteasome and p97 mediate mitophagy and degradation of mitofusins
induced by Parkin. Journal of Cell Biology 191, 1367-1380, doi:10.1083/jcb.201007013
(2010).
54 Chourasia, A. H., Boland, M. L. & Macleod, K. F. Mitophagy and cancer. Cancer and
Metabolism 3, 1-11, doi:10.1186/s40170-015-0130-8 (2015).
55 Saxton, W. M. & Hollenbeck, P. J. The axonal transport of mitochondria. Journal of Cell
Science 125, 2095-2104, doi:10.1242/jcs.053850 (2012).
56 O'Rourke, B. Mitochondrial ion channels. Annual Review of Physiology 69, 19-49,
doi:10.1146/annurev.physiol.69.031905.163804 (2007).
57 Giorgi, C., Marchi, S. & Pinton, P. The machineries, regulation and cellular functions of
mitochondrial calcium. Nature Reviews Molecular Cell Biology 19, 713-730,
doi:10.1038/s41580-018-0052-8 (2018).
58 Friel, D. D. & Tsien, R. W. An FCCP-sensitive Ca2+store in bullfrog sympathetic
neurons and its participation in stimulus-evoked changes in [Ca2+]i. Journal of
Neuroscience 14, 4007-4024, doi:10.1523/jneurosci.14-07-04007.1994 (1994).
59 Jean-Quartier, C. et al. Studying mitochondrial Ca 2+ uptake - A revisit. Molecular and
Cellular Endocrinology 353, 114-127, doi:10.1016/j.mce.2011.10.033 (2012).
60 Gunter, T. E. & Pfeiffer, D. R. Mechanisms by which mitochondria transport calcium.
American Journal of Physiology - Cell Physiology 258, C755-C786,
doi:10.1152/ajpcell.1990.258.5.c755 (1990).
120
61 Kamer, K. J. et al. MICU1 imparts the mitochondrial uniporter with the ability to
discriminate between Ca2+ and Mn2+. Proceedings of the National Academy of Sciences
of the United States of America 115, E7960-E7969, doi:10.1073/pnas.1807811115
(2018).
62 Granatiero, V., Pacifici, M., Raffaello, A., De Stefani, D. & Rizzuto, R. Overexpression
of Mitochondrial Calcium Uniporter Causes Neuronal Death. Oxidative Medicine and
Cellular Longevity 2019, doi:10.1155/2019/1681254 (2019).
63 Choi, S. et al. Mitochondrial calcium uniporter in Drosophila transfers calcium between
the endoplasmic reticulum and mitochondria in oxidative stress-induced cell death.
Journal of Biological Chemistry 292, 14473-14485, doi:10.1074/jbc.M116.765578
(2017).
64 Ren, T. et al. MCUR1-Mediated Mitochondrial Calcium Signaling Facilitates Cell
Survival of Hepatocellular Carcinoma via Reactive Oxygen Species-Dependent P53
Degradation. Antioxidants and Redox Signaling 28, 1120-1136,
doi:10.1089/ars.2017.6990 (2018).
65 O'Rourke, B., Cortassa, S. & Aon, M. A. Mitochondrial ion channels: Gatekeepers of life
and death. Physiology 20, 303-315, doi:10.1152/physiol.00020.2005 (2005).
66 Szewczyk, A., Jarmuszkiewicz, W. & Kunz, W. S. Mitochondrial potassium channels.
IUBMB Life 61, 134-143, doi:10.1002/iub.155 (2009).
67 Laskowski, M. et al. What do we not know about mitochondrial potassium channels?
Biochimica et Biophysica Acta - Bioenergetics 1857, 1247-1257,
doi:10.1016/j.bbabio.2016.03.007 (2016).
68 Bednarczyk, P. Potassium channels in brain mitochondria. Acta Biochimica Polonica 56,
385-392, doi:10.18388/abp.2009_2471 (2009).
69 Szabò, I., Leanza, L., Gulbins, E. & Zoratti, M. Physiology of potassium channels in the
inner membrane of mitochondria. Pflugers Archiv European Journal of Physiology 463,
231-246, doi:10.1007/s00424-011-1058-7 (2012).
121
70 Paggio, A. et al. Identification of an ATP-sensitive potassium channel in mitochondria.
Nature 572, 609-613, doi:10.1038/s41586-019-1498-3 (2019).
71 Rousset, S. et al. The Biology of Mitochondrial Uncoupling Proteins. Diabetes 53,
doi:10.2337/diabetes.53.2007.s130 (2004).
72 Ricquier, D. & Bouillaud, F. Mitochondrial uncoupling proteins: From mitochondria to
the regulation of energy balance. Journal of Physiology 529, 3-10, doi:10.1111/j.1469-
7793.2000.00003.x (2000).
73 Kwok, K. H. H. et al. Mitochondrial UCP5 is neuroprotective by preserving
mitochondrial membrane potential, ATP levels, and reducing oxidative stress in MPP+
and dopamine toxicity. Free Radical Biology and Medicine 49, 1023-1035,
doi:10.1016/j.freeradbiomed.2010.06.017 (2010).
74 Shimasaki, Y. et al. Uncoupling Protein 2 Impacts Endothelial Phenotype via p53-
Mediated Control of Mitochondrial Dynamics. Circulation Research 113, 891-901,
doi:10.1161/CIRCRESAHA.113.301319 (2013).
75 Colombini, M. VDAC: The channel at the interface between mitochondria and the
cytosol. Molecular and Cellular Biochemistry, 107-115 (2004).
76 Maldonado, E. N. & Lemasters, J. J. Warburg revisited: Regulation of mitochondrial
metabolism by voltage-dependent anion channels in cancer cells. Journal of
Pharmacology and Experimental Therapeutics 342, 637-641,
doi:10.1124/jpet.112.192153 (2012).
77 Szabo, I. & Zoratti, M. Mitochondrial channels: Ion fluxes and more. Physiological
Reviews 94, 519-608, doi:10.1152/physrev.00021.2013 (2014).
78 Premkumar, A. & Simantov, R. Mitochondrial voltage-dependent anion channel is
involved in dopamine-induced apoptosis. Journal of Neurochemistry 82, 345-352,
doi:10.1046/j.1471-4159.2002.00966.x (2002).
79 Sheldon, K. L., Maldonado, E. N., Lemasters, J. J., Rostovtseva, T. K. & Bezrukov, S. M.
Phosphorylation of voltage-dependent anion channel by serine/threonine kinases governs
122
its interaction with tubulin. PLoS ONE 6, 1-10, doi:10.1371/journal.pone.0025539
(2011).
80 Maldonado, E. N. et al. Voltage-dependent anion channels modulate mitochondrial
metabolism in cancer cells: regulation by free tubulin and erastin. J Biol Chem 288,
11920-11929, doi:10.1074/jbc.M112.433847 (2013).
81 Pastorino, J. G. & Hoek, J. B. Regulation of hexokinase binding to VDAC. Journal of
Bioenergetics and Biomembranes 40, 171-182, doi:10.1007/s10863-008-9148-8 (2008).
82 Dubey, A. K., Godbole, A. & Mathew, M. K. Regulation of VDAC trafficking modulates
cell death. Cell Death Discovery 2, doi:10.1038/cddiscovery.2016.85 (2016).
83 Fletcher, D. A. & Mullins, R. D. Cell mechanics and the cytoskeleton. Nature 463, 485-
492, doi:10.1038/nature08908 (2010).
84 Kuznetsov, A. V. et al. Crosstalk between Mitochondria and Cytoskeleton in Cardiac
Cells. Cells 9, 1-24 (2020).
85 Maldonado, E. N., Patnaik, J., Mullins, M. R. & Lemasters, J. J. Free tubulin modulates
mitochondrial membrane potential in cancer cells. Cancer Research 70, 10192-10201,
doi:10.1158/0008-5472.CAN-10-2429 (2010).
86 Kumazawa, A. et al. Microtubule disorganization affects the mitochondrial permeability
transition pore in cardiac myocytes. Circulation Journal 78, 1206-1215,
doi:10.1253/circj.CJ-13-1298 (2014).
87 Melli, G. et al. Alpha-lipoic acid prevents mitochondrial damage and neurotoxicity in
experimental chemotherapy neuropathy. Experimental Neurology 214, 276-284,
doi:10.1016/j.expneurol.2008.08.013 (2008).
88 Kandel, J., Angelin, A. A., Wallace, D. C. & Eckmann, D. M. Mitochondrial respiration
is sensitive to cytoarchitectural breakdown. Integrative Biology (United Kingdom) 8,
1170-1182, doi:10.1039/c6ib00192k (2016).
123
89 Rostovtseva, T. K. et al. Tubulin binding blocks mitochondrial voltage-dependent anion
channel and regulates respiration. Proceedings of the National Academy of Sciences of
the United States of America 105, 18746-18751, doi:10.1073/pnas.0806303105 (2008).
90 Boldogh, I. R. & Pon, L. A. Interactions of mitochondria with the actin cytoskeleton.
Biochim Biophys Acta 1763, 450-462, doi:10.1016/j.bbamcr.2006.02.014 (2006).
91 Davidson, A. J. & Wood, W. Unravelling the Actin Cytoskeleton: A New Competitive
Edge? Trends in Cell Biology 26, 569-576, doi:10.1016/j.tcb.2016.04.001 (2016).
92 Koya, R. C. et al. Gelsolin inhibits apoptosis by blocking mitochondrial membrane
potential loss and cytochrome c release. Journal of Biological Chemistry 275, 15343-
15349, doi:10.1074/jbc.275.20.15343 (2000).
93 Harms, C. et al. Neuronal gelsolin prevents apoptosis by enhancing actin
depolymerization. Molecular and Cellular Neuroscience 25, 69-82,
doi:10.1016/j.mcn.2003.09.012 (2004).
94 Gourlay, C. W. & Ayscough, K. R. The actin cytoskeleton: A key regulator of apoptosis
and ageing? Nature Reviews Molecular Cell Biology 6, 583-589, doi:10.1038/nrm1682
(2005).
95 Kusano, H. et al. Human gelsolin prevents apoptosis by inhibiting apoptotic
mitochondrial changes via closing VDAC. Oncogene 19, 4807-4814 (2000).
96 Foger, N., Rangell, L., Danilenko, D. M. & Chan, A. C. Requirement for Coronin 1 in T
Lymphocyte Trafficking and Cellular Homeostasis. Science 313, 839-842,
doi:10.1016/b978-1-4832-3292-8.50018-7 (2006).
97 Dustin, M. L. When F-actin becomes too much of a good thing. Science 313, 767-768,
doi:10.1126/science.1131714 (2006).
98 Moore, A. S., Wong, Y. C., Simpson, C. L. & Holzbaur, E. L. Dynamic actin cycling
through mitochondrial subpopulations locally regulates the fission-fusion balance within
mitochondrial networks. Nat Commun 7, 12886, doi:10.1038/ncomms12886 (2016).
124
99 Gourlay, C. W. & Ayscough, K. R. Identification of an upstream regulatory pathway
controlling actin-mediated apoptosis in yeast. Journal of Cell Science 118, 2119-2132,
doi:10.1242/jcs.02337 (2005).
100 Wang, Y. et al. Disruption of actin filaments induces mitochondrial Ca2+release to the
cytoplasm and [Ca2+]cchanges in Arabidopsis root hairs. BMC Plant Biology 10,
doi:10.1186/1471-2229-10-53 (2010).
101 Lo, Y. S. et al. Actin in mung bean mitochondria and implications for its function. Plant
Cell 23, 3727-3744, doi:10.1105/tpc.111.087403 (2011).
102 Matveeva, E. A., Venkova, L. S., Chernoivanenko, I. S. & Minin, A. A. Vimentin is
involved in regulation of mitochondrial motility and membrane potential by Rac1. Biol
Open 4, 1290-1297, doi:10.1242/bio.011874 (2015).
103 Nekrasova, O. E. et al. Vimentin intermediate filaments modulate the motility of
mitochondria. Molecular Biology of the Cell 22, 2282-2289, doi:10.1091/mbc.E10-09-
0766 (2011).
104 Chernoivanenko, I. S., Matveeva, E. A. & Minin, A. A. Vimentin intermediate filaments
increase mitochondrial membrane potential. Biochemistry (Moscow) Supplement Series
A: Membrane and Cell Biology 5, 21-28, doi:10.1134/S1990747811010041 (2011).
105 Chernoivanenko, I. S., Matveeva, E. A., Gelfand, V. I., Goldman, R. D. & Minin, A. A.
Mitochondrial membrane potential is regulated by vimentin intermediate filaments.
FASEB Journal 29, 820-827, doi:10.1096/fj.14-259903 (2015).
106 Hüttemann, M. et al. Regulation of oxidative phosphorylation, the mitochondrial
membrane potential, and their role in human disease. Journal of Bioenergetics and
Biomembranes 40, 445-456, doi:10.1007/s10863-008-9169-3 (2008).
107 Hüttemann, M., Lee, I., Samavati, L., Yu, H. & Doan, J. W. Regulation of mitochondrial
oxidative phosphorylation through cell signaling. Biochimica et Biophysica Acta -
Molecular Cell Research 1773, 1701-1720, doi:10.1016/j.bbamcr.2007.10.001 (2007).
125
108 Bijur, G. N. & Jope, R. S. Rapid accumulation of Akt in mitochondria following
phosphatidylinositol 3-kinase activation. J Neurochem 87, 1427-1435,
doi:10.1046/j.1471-4159.2003.02113.x (2003).
109 Lim, S. et al. Regulation of mitochondrial functions by protein phosphorylation and
dephosphorylation. Cell and Bioscience 6, 1-15, doi:10.1186/s13578-016-0089-3 (2016).
110 Balkwill, F. R., Capasso, M. & Hagemann, T. The tumor microenvironment at a glance. J
Cell Sci 125, 5591-5596, doi:10.1242/jcs.116392 (2012).
111 Whiteside, T. L. The tumor microenvironment and its role in promoting tumor growth.
Oncogene 27, 5904-5912, doi:10.1038/onc.2008.271 (2008).
112 Petrova, V., Annicchiarico-Petruzzelli, M., Melino, G. & Amelio, I. The hypoxic tumour
microenvironment. Oncogenesis 7, doi:10.1038/s41389-017-0011-9 (2018).
113 Ando, Y. et al. A Microdevice Platform Recapitulating Hypoxic Tumor
Microenvironments. Scientific Reports 7: 15233, 1-12, doi:10.1038/s41598-017-15583-3
(2017).
114 Winkler, J., Abisoye-Ogunniyan, A., Metcalf, K. J. & Werb, Z. Concepts of extracellular
matrix remodelling in tumour progression and metastasis. Nature Communications 11, 1-
19, doi:10.1038/s41467-020-18794-x (2020).
115 Chaudhuri, P. K., Low, B. C. & Lim, C. T. Mechanobiology of Tumor Growth. Chem
Rev 118, 6499-6515, doi:10.1021/acs.chemrev.8b00042 (2018).
116 Nia, H. T., Munn, L. L. & Jain, R. K. Physical traits of cancer. Science 370,
doi:10.1126/science.aaz0868 (2020).
117 Hawkins, B. J. et al. Mitochondrial complex II prevents hypoxic but not calcium- and
proapoptotic Bcl-2 protein-induced mitochondrial membrane potential loss. J Biol Chem
285, 26494-26505, doi:10.1074/jbc.M110.143164 (2010).
126
118 Wiese, M. et al. Hypoxia-mediated impairment of the mitochondrial respiratory chain
inhibits the bactericidal activity of macrophages. Infect Immun 80, 1455-1466,
doi:10.1128/IAI.05972-11 (2012).
119 Weinberg, J. M., Venkatachalam, M. A., Roeser, N. F. & Nissim, I. Mitochondrial
dysfunction during hypoxia/reoxygenation and its correction by anaerobic metabolism of
citric acid cycle intermediates. Proc Natl Acad Sci U S A 97, 2826-2831 (2000).
120 Khacho, M. et al. Acidosis overrides oxygen deprivation to maintain mitochondrial
function and cell survival. Nat Commun 5, 3550, doi:10.1038/ncomms4550 (2014).
121 Ma, Y. Y., Chen, H. W. & Tzeng, C. R. Low oxygen tension increases mitochondrial
membrane potential and enhances expression of antioxidant genes and implantation
protein of mouse blastocyst cultured in vitro. J Ovarian Res 10, 47, doi:10.1186/s13048-
017-0344-1 (2017).
122 Turcotte, M. L., Parliament, M., Franko, A. & Allalunis-Turner, J. Variation in
mitochondrial function in hypoxia-sensitive and hypoxia-tolerant human glioma cells.
British Journal of Cancer 86, 619-624, doi:10.1038/sj/bjc/6600087 (2002).
123 Papandreou, I., Cairns, R. A., Fontana, L., Lim, A. L. & Denko, N. C. HIF-1 mediates
adaptation to hypoxia by actively downregulating mitochondrial oxygen consumption.
Cell Metab 3, 187-197, doi:10.1016/j.cmet.2006.01.012 (2006).
124 Bertero, T. et al. Vascular stiffness mechanoactivates YAP/TAZ-dependent
glutaminolysis to drive pulmonary hypertension. J Clin Invest 126, 3313-3335,
doi:10.1172/JCI86387 (2016).
125 Morishima, M., Horikawa, K. & Funaki, M. Cardiomyocytes cultured on mechanically
compliant substrates, but not on conventional culture devices, exhibit prominent
mitochondrial dysfunction due to reactive oxygen species and insulin resistance under
high glucose. PLoS One 13, e0201891, doi:10.1371/journal.pone.0201891 (2018).
126 Bartolák-Suki, E., Imsirovic, J., Nishibori, Y., Krishnan, R. & Suki, B. Regulation of
mitochondrial structure and dynamics by the cytoskeleton and mechanical factors.
International Journal of Molecular Sciences 18, 7-11, doi:10.3390/ijms18081812 (2017).
127
127 Vaquero, E. C. et al. Extracellular matrix proteins protect pancreatic cancer cells from
death via mitochondrial and nonmitochondrial pathways. Gastroenterology 125, 1188-
1202, doi:10.1016/s0016-5085(03)01203-4 (2003).
128 Wu, X. et al. Pramlintide regulation of extracellular matrix (ECM) and apoptosis through
mitochondrial-dependent pathways in human nucleus pulposus cells. Int J Immunopathol
Pharmacol 31, 394632017747500, doi:10.1177/0394632017747500 (2018).
129 Jia, Y. et al. COMP-prohibitin 2 interaction maintains mitochondrial homeostasis and
controls smooth muscle cell identity. Cell Death Dis 9, 676, doi:10.1038/s41419-018-
0703-x (2018).
130 Jain, R. K., Martin, J. D. & Stylianopoulos, T. The role of mechanical forces in tumor
growth and therapy. Annu Rev Biomed Eng 16, 321-346, doi:10.1146/annurev-bioeng-
071813-105259 (2014).
131 Tse, J. M. et al. Mechanical compression drives cancer cells toward invasive phenotype.
Proc Natl Acad Sci U S A 109, 911-916, doi:10.1073/pnas.1118910109 (2012).
132 Kalli, M. et al. Mechanical Compression Regulates Brain Cancer Cell Migration Through
MEK1/Erk1 Pathway Activation and GDF15 Expression. Front Oncol 9, 992,
doi:10.3389/fonc.2019.00992 (2019).
133 Wang, F., Franco, R., Skotak, M., Hu, G. & Chandra, N. Mechanical stretch exacerbates
the cell death in SH-SY5Y cells exposed to paraquat: mitochondrial dysfunction and
oxidative stress. Neurotoxicology 41, 54-63, doi:10.1016/j.neuro.2014.01.002 (2014).
134 Liao, X. D., Wang, X. H., Jin, H. J., Y., C. L. & Chen, Q. Mechanical stretch induces
mitochondria-dependent apoptosis in neonatal rat cardiomyocytes and G2/M
accumulation in cardiac fibroblasts. Cell Research 14 (2004).
135 Bartolak-Suki, E. et al. Fluctuation-driven mechanotransduction regulates mitochondrial-
network structure and function. Nat Mater 14, 1049-1057, doi:10.1038/nmat4358 (2015).
136 Zielonka, J. et al. Mitochondria-Targeted Triphenylphosphonium-Based Compounds:
Syntheses, Mechanisms of Action, and Therapeutic and Diagnostic Applications. Chem
Rev 117, 10043-10120, doi:10.1021/acs.chemrev.7b00042 (2017).
128
137 Logan, A. et al. Assessing the mitochondrial membrane potential in cells and in vivo
using targeted click chemistry and mass spectrometry. Cell Metabolism 23, 379-385,
doi:10.1016/j.cmet.2015.11.014 (2016).
138 Perry, S. W., Norman, J. P., Barbieri, J., Brown, E. B. & Gelbard, H. A. Mitochondrial
membrane potential probes and the proton gradient: a practical usage guide.
Biotechniques 50, 98-115, doi:10.2144/000113610 (2011).
139 Martin, S. A., Hewish, M., Sims, D., Lord, C. J. & Ashworth, A. Parallel high-throughput
RNA interference screens identify PINK1 as a potential therapeutic target for the
treatment of DNA mismatch repair-deficient cancers. Cancer Res 71, 1836-1848,
doi:10.1158/0008-5472.CAN-10-2836 (2011).
140 Siegel, R. L., Miller, K. D. & Jemal, A. Cancer statistics, 2018. CA Cancer J Clin 68, 7-
30, doi:10.3322/caac.21442 (2018).
141 Morais, R. et al. Tumor-forming Ability in Athymic Nude Mice of Human Cell Lines
Devoid of Mitochondrial DNA. Cancer Research 54, 3889-3389 (1994).
142 Dong, L. F. et al. Horizontal transfer of whole mitochondria restores tumorigenic
potential in mitochondrial DNA-deficient cancer cells. Elife 6, doi:10.7554/eLife.22187
(2017).
143 Tan, A. S. et al. Mitochondrial genome acquisition restores respiratory function and
tumorigenic potential of cancer cells without mitochondrial DNA. Cell Metab 21, 81-94,
doi:10.1016/j.cmet.2014.12.003 (2015).
144 Dang, L. et al. Cancer-associated IDH1 mutations produce 2-hydroxyglutarate. Nature
462, 739-744, doi:10.1038/nature08617 (2009).
145 Sciacovelli, M. et al. Fumarate is an epigenetic modifier that elicits epithelial-to-
mesenchymal transition. Nature 537, 544-547, doi:10.1038/nature19353 (2016).
146 Yang, J. & Weinberg, R. A. Epithelial-mesenchymal transition: at the crossroads of
development and tumor metastasis. Dev Cell 14, 818-829,
doi:10.1016/j.devcel.2008.05.009 (2008).
129
147 Porporato, P. E., Filigheddu, N., Pedro, J. M. B., Kroemer, G. & Galluzzi, L.
Mitochondrial metabolism and cancer. Cell Res 28, 265-280, doi:10.1038/cr.2017.155
(2018).
148 Heerdt, B. G., Houston, M. A., Anthony, G. M. & Augenlicht, L. H. Mitochondrial
membrane potential in the coordination of p53-independent proliferation and apoptosis
pathways in human colonic carcinoma cells. Cancer Research 58, 2869-2875 (1998).
149 Kuwahara, Y. et al. The Involvement of Mitochondrial Membrane Potential in Cross-
Resistance Between Radiation and Docetaxel. International Journal of Radiation
Oncology*Biology*Physics 96, 556-565, doi:10.1016/j.ijrobp.2016.07.002 (2016).
150 Schieke, S. M. et al. Mitochondrial metabolism modulates differentiation and teratoma
formation capacity in mouse embryonic stem cells. J Biol Chem 283, 28506-28512,
doi:10.1074/jbc.M802763200 (2008).
151 LeBleu, V. S. et al. PGC-1alpha mediates mitochondrial biogenesis and oxidative
phosphorylation in cancer cells to promote metastasis. Nat Cell Biol 16, 992-1003, 1001-
1015, doi:10.1038/ncb3039 (2014).
152 Quail, D. F. & Joyce, J. A. Microenvironmental regulation of tumor progression and
metastasis. Nat Med 19, 1423-1437, doi:10.1038/nm.3394 (2013).
153 Cunha, G. R. Role of Mesenchymal-Epithelial Interactions in Normal and Abnormal
Development of the Mammary Gland and Prostate. Cancer 1030-1044 (1994).
154 Ronnov-Jessen, L., Petersen, O. W. & Bissell, M. J. Cellular Changes Involved in
Conversion of Normal to Malignant Breast: Importance of the Stromal Reaction.
Physiological Reviews 76, 69-125 (1996).
155 Shen, K. et al. Resolving cancer-stroma interfacial signalling and interventions with
micropatterned tumour-stromal assays. Nat Commun 5, 5662, doi:10.1038/ncomms6662
(2014).
156 Sgroi, D. C. Preinvasive breast cancer. Annu Rev Pathol 5, 193-221,
doi:10.1146/annurev.pathol.4.110807.092306 (2010).
130
157 Tajan, M. & Vousden, K. H. The Quid Pro Quo of the Tumor/Stromal Interaction. Cell
Metab 24, 645-646, doi:10.1016/j.cmet.2016.10.017 (2016).
158 Lyssiotis, C. A. & Kimmelman, A. C. Metabolic Interactions in the Tumor
Microenvironment. Trends Cell Biol 27, 863-875, doi:10.1016/j.tcb.2017.06.003 (2017).
159 Spill, F., Reynolds, D. S., Kamm, R. D. & Zaman, M. H. Impact of the physical
microenvironment on tumor progression and metastasis. Curr Opin Biotechnol 40, 41-48,
doi:10.1016/j.copbio.2016.02.007 (2016).
160 Stylianopoulos, T. et al. Causes, consequences, and remedies for growth-induced solid
stress in murine and human tumors. Proc Natl Acad Sci U S A 109, 15101-15108,
doi:10.1073/pnas.1213353109 (2012).
161 Carey, S. P., D'Alfonso, T. M., Shin, S. J. & Reinhart-King, C. A. Mechanobiology of
tumor invasion: engineering meets oncology. Crit Rev Oncol Hematol 83, 170-183,
doi:10.1016/j.critrevonc.2011.11.005 (2012).
162 Tomasek, J. J., Gabbiani, G., Hinz, B., Chaponnier, C. & Brown, R. A. Myofibroblasts
and mechano-regulation of connective tissue remodelling. Nat Rev Mol Cell Biol 3, 349-
363, doi:10.1038/nrm809 (2002).
163 Parekkadan, B. et al. Mesenchymal stem cell-derived molecules reverse fulminant
hepatic failure. PLoS One 2, e941, doi:10.1371/journal.pone.0000941 (2007).
164 Rajan, N., Habermehl, J., Cote, M. F., Doillon, C. J. & Mantovani, D. Preparation of
ready-to-use, storable and reconstituted type I collagen from rat tail tendon for tissue
engineering applications. Nat Protoc 1, 2753-2758, doi:10.1038/nprot.2006.430 (2006).
165 Komsa-Penkova, R., Spirova, R. & Bechev, B. Modification of Lowry’s method for
concentration measurement collagen. Journal of Biochemical and Biophysical Methods
32, 33-43 (1996).
166 Browne, A. W. et al. Structural and Functional Characterization of Human Stem-Cell-
Derived Retinal Organoids by Live Imaging. Invest Ophthalmol Vis Sci 58, 3311-3318,
doi:10.1167/iovs.16-20796 (2017).
131
167 Pertea, M., Kim, D., Pertea, G. M., Leek, J. T. & Salzberg, S. L. Transcript-level
expression analysis of RNA-seq experiments with HISAT, StringTie and Ballgown. Nat
Protoc 11, 1650-1667, doi:10.1038/nprot.2016.095 (2016).
168 Liao, Y., Smyth, G. K. & Shi, W. The Subread aligner: fast, accurate and scalable read
mapping by seed-and-vote. Nucleic Acids Res 41, e108, doi:10.1093/nar/gkt214 (2013).
169 Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion
for RNA-seq data with DESeq2. Genome Biol 15, 550, doi:10.1186/s13059-014-0550-8
(2014).
170 Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based approach for
interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 102, 15545-
15550, doi:10.1073/pnas.0506580102 (2005).
171 Vamsi K Mootha et al. PGC-1α-responsive genes involved in oxidative phosphorylation
are coordinately downregulated in human diabetes. Nature Genetics 34, 267-273 (2003).
172 Benjamini, Y. & Hochber, Y. Controlling the False Discovery Rate: a Practical and
Powerful Approach to Multiple Testing. J. R. Statist. Soc. B 57, 289-300 (1995).
173 Liberzon, A. et al. The Molecular Signatures Database (MSigDB) hallmark gene set
collection. Cell Syst 1, 417-425, doi:10.1016/j.cels.2015.12.004 (2015).
174 Liberzon, A. et al. Molecular signatures database (MSigDB) 3.0. Bioinformatics 27,
1739-1740, doi:10.1093/bioinformatics/btr260 (2011).
175 Lee, J., Wong, M., Smith, Q. & Baker, A. B. A novel system for studying mechanical
strain waveform-dependent responses in vascular smooth muscle cells. Lab Chip 13,
4573-4582, doi:10.1039/c3lc50894c (2013).
176 Iriondo, O. et al. TAK1 mediates microenvironment-triggered autocrine signals and
promotes triple-negative breast cancer lung metastasis. Nat Commun 9, 1994,
doi:10.1038/s41467-018-04460-w (2018).
132
177 Karnoub, A. E. et al. Mesenchymal stem cells within tumour stroma promote breast
cancer metastasis. Nature 449, 557-563, doi:10.1038/nature06188 (2007).
178 The Gene Ontology, C. Expansion of the Gene Ontology knowledgebase and resources.
Nucleic Acids Res 45, D331-D338, doi:10.1093/nar/gkw1108 (2017).
179 Ashburner, M. et al. Gene Ontology: tool for the unification of biology. Nature Genetics
25 (2000).
180 Kanehisa, M. & Goto, S. KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic
Acids Res 28 (2000).
181 Kanehisa, M., Sato, Y., Kawashima, M., Furumichi, M. & Tanabe, M. KEGG as a
reference resource for gene and protein annotation. Nucleic Acids Res 44, D457-462,
doi:10.1093/nar/gkv1070 (2016).
182 Kanehisa, M., Furumichi, M., Tanabe, M., Sato, Y. & Morishima, K. KEGG: new
perspectives on genomes, pathways, diseases and drugs. Nucleic Acids Res 45, D353-
D361, doi:10.1093/nar/gkw1092 (2017).
183 Wei, T. Y. et al. Protein arginine methyltransferase 5 is a potential oncoprotein that
upregulates G1 cyclins/cyclin-dependent kinases and the phosphoinositide 3-kinase/AKT
signaling cascade. Cancer Sci 103, 1640-1650, doi:10.1111/j.1349-7006.2012.02367.x
(2012).
184 Wang, Y. et al. Gene-expression profiles to predict distant metastasis of lymph-node-
negative primary breast cancer. The Lancet 365, 671-679, doi:10.1016/s0140-
6736(05)70933-8 (2005).
185 Anastassiou, D. et al. Human cancer cells express Slug-based epithelialmesenchymal
transition gene expression signature obtained in vivo. BMC Cancer 11 (2011).
186 Cunningham, J. T. et al. mTOR controls mitochondrial oxidative function through a
YY1-PGC-1alpha transcriptional complex. Nature 450, 736-740,
doi:10.1038/nature06322 (2007).
133
187 Scaduto, R. C. & Grotyohann, L. W. Measurement of Mitochondrial Membrane Potential
Using Fluorescent Rhodamine Derivatives. Biophysical Journal 76, 467-477 (1999).
188 Burbulla, L. F. et al. Mitochondrial proteolytic stress induced by loss of mortalin function
is rescued by Parkin and PINK1. Cell Death Dis 5, e1180, doi:10.1038/cddis.2014.103
(2014).
189 De Ruyck, J. et al. Towards the understanding of the absorption spectra of
NAD(P)H/NAD(P)+ as a common indicator of dehydrogenase enzymatic activity.
Chemical Physics Letters 450, 119-122, doi:10.1016/j.cplett.2007.10.092 (2007).
190 Blacker, T. S. & Duchen, M. R. Investigating mitochondrial redox state using NADH and
NADPH autofluorescence. Free Radic Biol Med 100, 53-65,
doi:10.1016/j.freeradbiomed.2016.08.010 (2016).
191 Alhallak, K., Rebello, L. G., Muldoon, T. J., Quinn, K. P. & Rajaram, N. Optical redox
ratio identifies metastatic potential-dependent changes in breast cancer cell metabolism.
Biomed Opt Express 7, 4364-4374, doi:10.1364/BOE.7.004364 (2016).
192 Dumollard, R., Ward, Z., Carroll, J. & Duchen, M. R. Regulation of redox metabolism in
the mouse oocyte and embryo. Development 134, 455-465, doi:10.1242/dev.02744
(2007).
193 Huang, J., Wu, S., Barrera, J., Matthews, K. & Pan, D. The Hippo signaling pathway
coordinately regulates cell proliferation and apoptosis by inactivating Yorkie, the
Drosophila Homolog of YAP. Cell 122, 421-434, doi:10.1016/j.cell.2005.06.007 (2005).
194 Pan, D. The hippo signaling pathway in development and cancer. Dev Cell 19, 491-505,
doi:10.1016/j.devcel.2010.09.011 (2010).
195 Dupont, S. et al. Role of YAP/TAZ in mechanotransduction. Nature 474, 179-183,
doi:10.1038/nature10137 (2011).
196 Das, A., Fischer, R. S., Pan, D. & Waterman, C. M. YAP Nuclear Localization in the
Absence of Cell-Cell Contact Is Mediated by a Filamentous Actin-dependent, Myosin II-
and Phospho-YAP-independent Pathway during Extracellular Matrix Mechanosensing. J
Biol Chem 291, 6096-6110, doi:10.1074/jbc.M115.708313 (2016).
134
197 Monod, J. The Growth of Bacterial Cultures. Annual Reviews Microbiology 3, 371-394
(1949).
198 Zhao, B. et al. Inactivation of YAP oncoprotein by the Hippo pathway is involved in cell
contact inhibition and tissue growth control. Genes Dev 21, 2747-2761,
doi:10.1101/gad.1602907 (2007).
199 Svoboda, K. K., Moessner, P., Field, T. & Acevedo, J. ROCK inhibitor (Y27632)
increases apoptosis and disrupts the actin cortical mat in embryonic avian corneal
epithelium. Dev Dyn 229, 579-590, doi:10.1002/dvdy.20008 (2004).
200 Coué, M., Brenner, S. L., Spector, I. & Korn, E. D. Inhibition of actin polymerization by
latrunculin A. FEBS Letters 213 (1987).
201 Furukawa, K. T., Yamashita, K., Sakurai, N. & Ohno, S. The Epithelial Circumferential
Actin Belt Regulates YAP/TAZ through Nucleocytoplasmic Shuttling of Merlin. Cell
Rep 20, 1435-1447, doi:10.1016/j.celrep.2017.07.032 (2017).
202 Omelchenko, T., Vasiliev, J. M., Gelfand, I. M., Feder, H. H. & Bonder, E. M.
Mechanisms of polarization of the shape of fibroblasts and epitheliocytes: Separation of
the roles of microtubules and Rho-dependent actin-myosin contractility. Proc Natl Acad
Sci U S A 99, 10452-10457, doi:10.1073/pnas.152339899 (2002).
203 Harris, A. R., Jreij, P. & Fletcher, D. A. Mechanotransduction by the Actin Cytoskeleton:
Converting Mechanical Stimuli into Biochemical Signals. Annual Review of Biophysics
47, 617-631, doi:10.1146/annurev- (2018).
204 Ohashi, K., Fujiwara, S. & Mizuno, K. Roles of the cytoskeleton, cell adhesion and rho
signalling in mechanosensing and mechanotransduction. J Biochem 161, 245-254,
doi:10.1093/jb/mvw082 (2017).
205 Karmali, P. P. et al. Metastasis of tumor cells is enhanced by downregulation of Bit1.
PLoS One 6, e23840, doi:10.1371/journal.pone.0023840 (2011).
206 Barkan, D. et al. Inhibition of metastatic outgrowth from single dormant tumor cells by
targeting the cytoskeleton. Cancer Res 68, 6241-6250, doi:10.1158/0008-5472.CAN-07-
6849 (2008).
135
207 Xu, H. N., Tchou, J. & Li, L. Z. Redox imaging of human breast cancer core biopsies: a
preliminary investigation. Acad Radiol 20, 764-768, doi:10.1016/j.acra.2013.02.006
(2013).
208 Xu, H. N., Tchou, J., Feng, M., Zhao, H. & Li, L. Z. Optical redox imaging indices
discriminate human breast cancer from normal tissues. J Biomed Opt 21, 114003,
doi:10.1117/1.JBO.21.11.114003 (2016).
209 Li, L. Z. et al. Quantitative magnetic resonance and optical imaging biomarkers of
melanoma metastatic potential. Proc Natl Acad Sci U S A 106, 6608-6613,
doi:10.1073/pnas.0901807106 (2009).
210 Ralph, S. J., Rodríguez-Enríquez, S., Neuzil, J., Saavedra, E. & Moreno-Sánchez, R. The
causes of cancer revisited: “Mitochondrial malignancy” and ROS-induced oncogenic
transformation – Why mitochondria are targets for cancer therapy. Molecular Aspects of
Medicine 31, 145-170, doi:10.1016/j.mam.2010.02.008 (2010).
211 Martinez-Outschoorn, U. E. et al. Stromal-epithelial metabolic coupling in cancer:
integrating autophagy and metabolism in the tumor microenvironment. Int J Biochem
Cell Biol 43, 1045-1051, doi:10.1016/j.biocel.2011.01.023 (2011).
212 Morris, B. A. et al. Collagen Matrix Density Drives the Metabolic Shift in Breast Cancer
Cells. EBioMedicine 13, 146-156, doi:10.1016/j.ebiom.2016.10.012 (2016).
213 Scharping, N. E. et al. The Tumor Microenvironment Represses T Cell Mitochondrial
Biogenesis to Drive Intratumoral T Cell Metabolic Insufficiency and Dysfunction.
Immunity 45, 374-388, doi:10.1016/j.immuni.2016.07.009 (2016).
214 van Brussel, A. S. et al. Hypoxia-Targeting Fluorescent Nanobodies for Optical
Molecular Imaging of Pre-Invasive Breast Cancer. Mol Imaging Biol 18, 535-544,
doi:10.1007/s11307-015-0909-6 (2016).
215 Malandrino, A., Mak, M., Kamm, R. D. & Moeendarbary, E. Complex mechanics of the
heterogeneous extracellular matrix in cancer. Extreme Mech Lett 21, 25-34,
doi:10.1016/j.eml.2018.02.003 (2018).
136
216 Pickup, M. W., Mouw, J. K. & Weaver, V. M. The extracellular matrix modulates the
hallmarks of cancer. EMBO Rep 15, 1243-1253, doi:10.15252/embr.201439246 (2014).
217 DelNero, P., Hopkins, B. D., Cantley, L. C. & Fischbach, C. Cancer metabolism gets
physical. Science Translational Medicine 10 (2018).
218 Hu, H. et al. Phosphoinositide 3-Kinase Regulates Glycolysis through Mobilization of
Aldolase from the Actin Cytoskeleton. Cell 164, 433-446, doi:10.1016/j.cell.2015.12.042
(2016).
219 Chan, Y.-H. M. & Marshall, W. F. Scaling properties of cell and organelle size.
Organogenesis 6, 88-96, doi:10.4161/org.6.2.11464 (2010).
220 Miettinen, T. P. & Bjorklund, M. Cellular Allometry of Mitochondrial Functionality
Establishes the Optimal Cell Size. Dev Cell 39, 370-382,
doi:10.1016/j.devcel.2016.09.004 (2016).
221 Rafelski, S. M. et al. Mitochondrial network size scaling in budding yeast. Science 338,
822-824, doi:10.1126/science.1225720 (2012).
222 Nagaraj, R. et al. Control of mitochondrial structure and function by the Yorkie/YAP
oncogenic pathway. Genes Dev 26, 2027-2037, doi:10.1101/gad.183061.111 (2012).
223 von Eyss, B. et al. A MYC-Driven Change in Mitochondrial Dynamics Limits YAP/TAZ
Function in Mammary Epithelial Cells and Breast Cancer. Cancer Cell 28, 743-757,
doi:10.1016/j.ccell.2015.10.013 (2015).
224 Xie, X., Venit, T., Drou, N. & Percipalle, P. In Mitochondria ?-Actin Regulates mtDNA
Transcription and Is Required for Mitochondrial Quality Control. iScience 3, 226-237,
doi:10.1016/j.isci.2018.04.021 (2018).
225 Porazinski, S. et al. YAP is essential for tissue tension to ensure vertebrate 3D body
shape. Nature 521, 217-221, doi:10.1038/nature14215 (2015).
226 Qiao, Y. et al. YAP Regulates Actin Dynamics through ARHGAP29 and Promotes
Metastasis. Cell Rep 19, 1495-1502, doi:10.1016/j.celrep.2017.04.075 (2017).
137
227 Amano, M., Nakayama, M. & Kaibuchi, K. Rho-kinase/ROCK: A key regulator of the
cytoskeleton and cell polarity. Cytoskeleton (Hoboken) 67, 545-554,
doi:10.1002/cm.20472 (2010).
228 Fiaschi, T. et al. Reciprocal metabolic reprogramming through lactate shuttle
coordinately influences tumor-stroma interplay. Cancer Res 72, 5130-5140,
doi:10.1158/0008-5472.CAN-12-1949 (2012).
229 Sonveaux, P. et al. Targeting lactate-fueled respiration selectively kills hypoxic tumor
cells in mice. J Clin Invest 118, 3930-3942, doi:10.1172/JCI36843 (2008).
230 Gao, D., Rahbar, R. & Fish, E. N. CCL5 activation of CCR5 regulates cell metabolism to
enhance proliferation of breast cancer cells. Open Biol 6, doi:10.1098/rsob.160122
(2016).
231 D A Bronzert et al. Synthesis and secretion of platelet-derived growth factor by human
breastcancercelllines. Proc Natl Acad Sci U S A 84, 5763-5767 (1987).
232 Tobar, N. et al. Soluble MMP-14 produced by bone marrow-derived stromal cells sheds
epithelial endoglin modulating the migratory properties of human breast cancer cells.
Carcinogenesis 35, 1770-1779, doi:10.1093/carcin/bgu061 (2014).
233 Haslam, S. Z., Counterman, L. J. & Nummy, K. A. in Biology of the Cancer Cell
Advances in Molecular and Cell Biology 115-130 (1993).
234 Kalluri, R. The biology and function of fibroblasts in cancer. Nat Rev Cancer 16, 582-
598, doi:10.1038/nrc.2016.73 (2016).
235 Welsh, J. in Animal Models for the Study of Human Disease 997-1018 (2013).
236 Ye, X.-Q. et al. Mitochondrial and energy metabolism-related properties as novel
indicators of lung cancer stem cells. International Journal of Cancer 129, 820-831,
doi:10.1002/ijc.25944 (2011).
138
237 Begum, H. M. et al. Spatial Regulation of Mitochondrial Heterogeneity by Stromal
Confinement in Micropatterned Tumor Models. Sci Rep 9, 11187, doi:10.1038/s41598-
019-47593-8 (2019).
238 Anderson, N. M. & Simon, M. C. The tumor microenvironment. Curr Biol 30, R921-
R925, doi:10.1016/j.cub.2020.06.081 (2020).
239 Carmona-Fontaine, C. et al. Metabolic origins of spatial organization in the tumor
microenvironment. Proc Natl Acad Sci U S A 114, 2934-2939,
doi:10.1073/pnas.1700600114 (2017).
240 Ando, Y. et al. A Microdevice Platform Recapitulating Hypoxic Tumor
Microenvironments. Sci Rep 7, 15233, doi:10.1038/s41598-017-15583-3 (2017).
241 Pietras, K. & Ostman, A. Hallmarks of cancer: interactions with the tumor stroma. Exp
Cell Res 316, 1324-1331, doi:10.1016/j.yexcr.2010.02.045 (2010).
242 Federica Sotgia et al. Understanding the Warburg effect and the prognostic value of
stromal caveolin-1 as a marker of a lethal tumor microenvironment. Breast Cancer Res
13 (2011).
243 Paul, C. D., Hung, W. C., Wirtz, D. & Konstantopoulos, K. Engineered Models of
Confined Cell Migration. Annu Rev Biomed Eng 18, 159-180, doi:10.1146/annurev-
bioeng-071114-040654 (2016).
244 Balzer, E. M. et al. Physical confinement alters tumor cell adhesion and migration
phenotypes. FASEB J 26, 4045-4056, doi:10.1096/fj.12-211441 (2012).
245 Harris, T. J. & Tepass, U. Adherens junctions: from molecules to morphogenesis. Nat
Rev Mol Cell Biol 11, 502-514, doi:10.1038/nrm2927 (2010).
246 Onder, T. T. et al. Loss of E-cadherin promotes metastasis via multiple downstream
transcriptional pathways. Cancer Res 68, 3645-3654, doi:10.1158/0008-5472.CAN-07-
2938 (2008).
139
247 Na, T. Y., Schecterson, L., Mendonsa, A. M. & Gumbiner, B. M. The functional activity
of E-cadherin controls tumor cell metastasis at multiple steps. Proc Natl Acad Sci U S A
117, 5931-5937, doi:10.1073/pnas.1918167117 (2020).
248 Padmanaban, V. et al. E-cadherin is required for metastasis in multiple models of breast
cancer. Nature 573, 439-444, doi:10.1038/s41586-019-1526-3 (2019).
249 Miranda, K. C. et al. A dileucine motif targets E-cadherin to the basolateral cell surface
in Madin-Darby canine kidney and LLC-PK1 epithelial cells. J Biol Chem 276, 22565-
22572, doi:10.1074/jbc.M101907200 (2001).
250 Nelly Auersperg et al. E-cadherin induces mesenchymal-to-epithelial transition in human
ovarian surface epithelium. Proc Natl Acad Sci U S A 96, 6249–6254 (1999).
251 Hruz, T. et al. Genevestigator v3: a reference expression database for the meta-analysis
of transcriptomes. Adv Bioinformatics 2008, 420747, doi:10.1155/2008/420747 (2008).
252 Ozawa, M., Ringwald, M. & Kemler, R. Uvomorulin-catenin complex formation is
regulated by a specific domain in the cytoplasmic region of the cell adhesion molecule.
Proc Natl Acad Sci U S A 87, 4246-4250 (1990).
253 Bruckner, B. R. & Janshoff, A. Importance of integrity of cell-cell junctions for the
mechanics of confluent MDCK II cells. Sci Rep 8, 14117, doi:10.1038/s41598-018-
32421-2 (2018).
254 Gross, J. C., Schreiner, A., Engels, K. & Starzinski-Powitz, A. E-cadherin Surface Levels
in Epithelial Growth Factor-stimulated Cells Depend on Adherens Junction Protein
Shrew-1. Molecular Biology of the Cell 20, 3598-3607, doi:10.1091/mbc.e08-12-1240
(2009).
255 Xie, G. et al. E-Cadherin-Mediated Cell Contact Controls the Epidermal Damage
Response in Radiation Dermatitis. J Invest Dermatol 137, 1731-1739,
doi:10.1016/j.jid.2017.03.036 (2017).
256 Lim, S. O. et al. Epigenetic changes induced by reactive oxygen species in hepatocellular
carcinoma: methylation of the E-cadherin promoter. Gastroenterology 135, 2128-2140,
2140 e2121-2128, doi:10.1053/j.gastro.2008.07.027 (2008).
140
257 Park, S. Y., Shin, J.-H. & Kee, S.-H. E-cadherin expression increases cell proliferation by
regulating energy metabolism through nuclear factor-κB in AGS cells. Cancer Science
108, 1769-1777, doi:10.1111/cas.13321 (2017).
258 Sing, A. et al. The atypical cadherin fat directly regulates mitochondrial function and
metabolic state. Cell 158, 1293-1308, doi:10.1016/j.cell.2014.07.036 (2014).
259 Liu, H. et al. Multifaceted regulation and functions of YAP/TAZ in tumors (Review).
Oncol Rep 40, 16-28, doi:10.3892/or.2018.6423 (2018).
260 Warren, J. S. A., Xiao, Y. & Lamar, J. M. YAP/TAZ Activation as a Target for Treating
Metastatic Cancer. Cancers (Basel) 10, doi:10.3390/cancers10040115 (2018).
261 Overholtzer, M. et al. Transforming properties of YAP, a candidate oncogene on the
chromosome 11q22 amplicon. Proc Natl Acad Sci U S A 103, 12405-12410 (2006).
262 Cordenonsi, M. et al. The Hippo transducer TAZ confers cancer stem cell-related traits
on breast cancer cells. Cell 147, 759-772, doi:10.1016/j.cell.2011.09.048 (2011).
263 Kim, H. M., H., J. W. & Koo, J. S. Expression of Yes-associated protein (YAP) in
metastatic breast cancer. Int J Clin Exp Pathol. 8, 11248–11257 (2015).
264 Kim, N. G., Koh, E., Chen, X. & Gumbiner, B. M. E-cadherin mediates contact inhibition
of proliferation through Hippo signaling-pathway components. Proc Natl Acad Sci U S A
108, 11930-11935, doi:10.1073/pnas.1103345108 (2011).
265 Huang, S. et al. Yap regulates mitochondrial structural remodeling during myoblast
differentiation. Am J Physiol Cell Physiol 315, C474-C484,
doi:10.1152/ajpcell.00112.2018 (2018).
266 Zhang, X., Li, F., Cui, Y., Liu, S. & Sun, H. Mst1 overexpression combined with Yap
knockdown augments thyroid carcinoma apoptosis via promoting MIEF1-related
mitochondrial fission and activating the JNK pathway. Cancer Cell Int 19, 143,
doi:10.1186/s12935-019-0860-8 (2019).
141
267 Yan, H. et al. Yap regulates gastric cancer survival and migration via
SIRT1/Mfn2/mitophagy. Oncol Rep 39, 1671-1681, doi:10.3892/or.2018.6252 (2018).
268 Zhou, Y. et al. YAP promotes multi-drug resistance and inhibits autophagy-related cell
death in hepatocellular carcinoma via the RAC1-ROS-mTOR pathway. Cancer Cell
International 19, doi:10.1186/s12935-019-0898-7 (2019).
269 Yuan, L. et al. Palmitic acid dysregulates the Hippo-YAP pathway and inhibits
angiogenesis by inducing mitochondrial damage and activating the cytosolic DNA sensor
cGAS-STING-IRF3 signaling mechanism. J Biol Chem 292, 15002-15015,
doi:10.1074/jbc.M117.804005 (2017).
270 Miyazono, Y. et al. Uncoupled mitochondria quickly shorten along their long axis to
form indented spheroids, instead of rings, in a fission-independent manner. Sci Rep 8,
350, doi:10.1038/s41598-017-18582-6 (2018).
271 Fan, S. et al. PINK1-Dependent Mitophagy Regulates the Migration and Homing of
Multiple Myeloma Cells via the MOB1B-Mediated Hippo-YAP/TAZ Pathway. Adv Sci
(Weinh) 7, 1900860, doi:10.1002/advs.201900860 (2020).
272 Basu, S., Totty, N. F., Irwin, M. S., Sudol, M. & Downward, J. Akt Phosphorylates the
Yes-Associated Protein, YAP, to Induce Interaction with 14-3-3 and Attenuation of p73-
Mediated Apoptosis. Molecular Cell 11, 11-23, doi:10.1016/s1097-2765(02)00776-1
(2003).
273 Robinson, B. S. & Moberg, K. H. Cell-cell junctions: alpha-catenin and E-cadherin help
fence in Yap1. Curr Biol 21, R890-892, doi:10.1016/j.cub.2011.09.019 (2011).
274 Zhao, B., Li, L., Tumaneng, K., Wang, C. Y. & Guan, K. L. A coordinated
phosphorylation by Lats and CK1 regulates YAP stability through SCF(beta-TRCP).
Genes Dev 24, 72-85, doi:10.1101/gad.1843810 (2010).
275 Wang, W. et al. AMPK modulates Hippo pathway activity to regulate energy
homeostasis. Nat Cell Biol 17, 490-499, doi:10.1038/ncb3113 (2015).
142
276 Schneider, A. et al. Single organelle analysis to characterize mitochondrial function and
crosstalk during viral infection. Sci Rep 9, 8492, doi:10.1038/s41598-019-44922-9
(2019).
277 Liu-Chittenden, Y. et al. Genetic and pharmacological disruption of the TEAD-YAP
complex suppresses the oncogenic activity of YAP. Genes Dev 26, 1300-1305,
doi:10.1101/gad.192856.112 (2012).
278 Wang, C. et al. Verteporfin inhibits YAP function through up-regulating 14-3-3σ
sequestering YAP in the cytoplasm. Am J Cancer Res 6, 27-37 (2016).
279 Kastan, N. et al. Small-molecule inhibition of Lats kinases may promote Yap-dependent
proliferation in postmitotic mammalian tissues. Nat Commun 12, 3100,
doi:10.1038/s41467-021-23395-3 (2021).
280 Bertolo, A., Baur, M., Guerrero, J., Potzel, T. & Stoyanov, J. Autofluorescence is a
Reliable in vitro Marker of Cellular Senescence in Human Mesenchymal Stromal Cells.
Sci Rep 9, 2074, doi:10.1038/s41598-019-38546-2 (2019).
281 Roedder, S. et al. Expression of Mitochondrial-Encoded Genes in Blood Differentiate
Acute Renal Allograft Rejection. Front Med (Lausanne) 4, 185,
doi:10.3389/fmed.2017.00185 (2017).
282 Gureev, A. P., Shaforostova, E. A. & Popov, V. N. Regulation of Mitochondrial
Biogenesis as a Way for Active Longevity: Interaction Between the Nrf2 and PGC-
1alpha Signaling Pathways. Front Genet 10, 435, doi:10.3389/fgene.2019.00435 (2019).
283 McFate, T. et al. Pyruvate dehydrogenase complex activity controls metabolic and
malignant phenotype in cancer cells. J Biol Chem 283, 22700-22708,
doi:10.1074/jbc.M801765200 (2008).
284 Sohn, Y. S. et al. NAF-1 and mitoNEET are central to human breast cancer proliferation
by maintaining mitochondrial homeostasis and promoting tumor growth. Proc Natl Acad
Sci U S A 110, 14676-14681, doi:10.1073/pnas.1313198110 (2013).
285 Pocaterra, A., Romani, P. & Dupont, S. YAP/TAZ functions and their regulation at a
glance. Journal of Cell Science 133, jcs230425, doi:10.1242/jcs.230425 (2020).
143
286 Zhang, X. et al. The role of YAP/TAZ activity in cancer metabolic reprogramming.
Molecular Cancer 17, doi:10.1186/s12943-018-0882-1 (2018).
287 Dupont, S. Role of YAP/TAZ in cell-matrix adhesion-mediated signalling and
mechanotransduction. Exp Cell Res 343, 42-53, doi:10.1016/j.yexcr.2015.10.034 (2016).
288 Enzo, E. et al. Aerobic glycolysis tunes YAP/TAZ transcriptional activity. EMBO J 34,
1349-1370, doi:10.15252/embj.201490379 (2015).
289 Kim, M., Kim, T., Johnson, R. L. & Lim, D. S. Transcriptional co-repressor function of
the hippo pathway transducers YAP and TAZ. Cell Rep 11, 270-282,
doi:10.1016/j.celrep.2015.03.015 (2015).
290 Liu, Y. et al. CCT3 acts upstream of YAP and TFCP2 as a potential target and tumour
biomarker in liver cancer. Cell Death Dis 10, 644, doi:10.1038/s41419-019-1894-5
(2019).
291 White, S. M. et al. YAP/TAZ Inhibition Induces Metabolic and Signaling Rewiring
Resulting in Targetable Vulnerabilities in NF2-Deficient Tumor Cells. Dev Cell 49, 425-
443 e429, doi:10.1016/j.devcel.2019.04.014 (2019).
Abstract (if available)
Abstract
Cancer cells have an abnormally high mitochondrial membrane potential (ΔΨm), which is associated with enhanced invasive properties in vitro. The mechanisms underlying the abnormal ΔΨm in cancer cells remain unclear. Research on different cell types has shown that ΔΨm is regulated by various intracellular mechanisms such as by mitochondrial inner and outer membrane ion transporters, cytoskeletal elements and biochemical signaling pathways. On the other hand, the role of extrinsic, tumor microenvironment (TME) derived cues in regulating ΔΨm is not well defined. In this dissertation, we first summarize the existing literature on intercellular mechanisms of ΔΨm regulation, with a focus on cancer cells. We then offer our perspective on the different ways through which the microenvironmental cues such as mechanical stresses may regulate cancer cell ΔΨm. In the subsequent chapters we present our work on recapitulating mitochondrial heterogeneity in vitro using a micropatterning based approach and the mechanisms that regulate the spatial gradients of ΔΨm observed in them.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Membrane-bound regulation of hematopoietic stem cells
PDF
Context-dependent role of androgen receptor (AR) in estrogen receptor-positive (ER+) breast cancer
PDF
Roles of circadian clock genes in cancer
PDF
Development of a colorectal cancer-on-chip to investigate the tumor microenvironment's role in cancer progression
PDF
The intersection of mitochondrial biology and cancer: insights from mitochondrial microproteins and mtDNA alterations
PDF
Insights from a computational approach to natural killer cell activation
PDF
Understanding prostate cancer genetic susceptibility and chromatin regulation
PDF
Genome-scale modeling of macrophage activity in the colorectal cancer microenvironment
PDF
Biophysical studies of nanosecond pulsed electric field induced cell membrane permeabilization
PDF
Investigating the complexity of the tumor microenvironment's role in drug response
PDF
Ectopic expression of a truncated isoform of hair keratin 81 in breast cancer alters biophysical characteristics to promote metastatic propensity
PDF
Mitochondrial dynamics regulate Leydig cell health and integrity
PDF
Engineering immunotoxin and viral vectors for cancer therapy
PDF
Extracellular matrix regulation of mitochondrial function in engineered cardiac myocytes
PDF
Using chemical biology approaches to investigate the consequences of protein concentration and activity in cancer cells
PDF
Contribution of cancer associated fibroblasts to cancer progression
PDF
Molecularly targeted micelle nanoparticles for cancer drug delivery and lymph node metastasis detection
PDF
Mechanistic modeling of angiogenic factors network and cancer therapy
PDF
An image-based identification of aggressive circulating tumor cell subtypes
PDF
Role of a novel transmembrane protein, MTTS1 in mitochondrial regulation and tumor suppression
Asset Metadata
Creator
Begum, Hydari Masuma
(author)
Core Title
Microenvironmental and biomechanical regulation of mitochondrial membrane potential in cancer cells
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Biomedical Engineering
Degree Conferral Date
2022-05
Publication Date
04/18/2024
Defense Date
03/01/2022
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
adherens junction,breast cancer,cancer metabolism,CRISPR/Cas9,E-cadherin,MCF-7,MDA-MB-231,mitochondrial membrane potential,OAI-PMH Harvest,tumor microenvironment,YAP/TAZ
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Shen, Keyue (
committee chair
), Finley, Stacey (
committee member
), Khoo, Michael (
committee member
), Mumenthaler, Shannon (
committee member
), Yu, Min (
committee member
)
Creator Email
hbegum@usc.edu,hmb.4494@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-oUC111014554
Unique identifier
UC111014554
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Begum, Hydari Masuma
Type
texts
Source
20220419-usctheses-batch-929
(batch),
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Repository Email
cisadmin@lib.usc.edu
Tags
adherens junction
breast cancer
cancer metabolism
CRISPR/Cas9
E-cadherin
MCF-7
MDA-MB-231
mitochondrial membrane potential
tumor microenvironment
YAP/TAZ