Characterization of the Cardiogenesis of Embryonic Stem Cells
by
Chen-rei Wan
MASSACHUSETTS INSTITUTE
OF TECHNOLOGY
M.S. Mechanical Engineering
Massachusetts Institute of Technology, 2005
MAY 18 2011
LIBRARIES
B.S. Mechanical Engineering
Carnegie Mellon University, 2001
ARCHIVES
Submitted to the Department of Mechanical Engineering
in Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy in Mechanical Engineering
at the
Massachusetts Institute of Technology
February 2011
C 2010 Massachusetts Institute of Technology. All rights reserved
Signature of the Author:
Certified by:
Roger D. Kamm, Ph.D
Singapore Research Professor of Biological and Mechanical Engineering
Thesis Chair and Supervisor
Accepted by:
David Hardt, Ph.D
Chairman, Department Committee on Graduate Students
Characterization of the Cardiogenesis of Embryonic Stem Cells
by
Chen-rei Wan
Submitted to the Department of Mechanical Engineering
on 12/06/2010 in fulfillment of the
requirements for the Degree of PhD in
Mechanical Engineering
Abstract
Cardiovascular diseases persist as the leading cause of mortality worldwide. Stem cell therapy,
aimed to restore contractility and proper vasculature, has gained considerable attention as an
attractive therapeutic option. However, proper cell differentiation, survival and integration in an
infarcted zone remain elusive. This thesis aims to utilize in vitro techniques to obtain a
systematic characterization of how individual stimulations can affect the cardiogenesis process
of embryonic stem cells.
First, a compliant microfluidic system was developed to study the individual and combined
effects of culture dimensions and uniaxial cyclic stretch on the differentiation process. A smaller
culture dimension, with a characteristic length scale of hundreds of micrometers, dramatically
enhanced differentiation partly due to an accumulation of cell-secreted and cardiogenic BMP2.
Uniaxial cyclic stretch, on the other hand, inhibited differentiation. With this microfluidic
platform and a GFP-reporting differentiation cell line, effects of various external stimuli on
differentiation were systematically studied.
Next, the effects of collagen I and cell alignment, two biophysical signatures of the adult
myocardium, on promoting phenotypic changes of isolated embryonic stem cell derived
cardiomyocytes (ESCDMs) were investigated. Effects of collagen I depended on how it was
presented to the cells and overlaying collagen gel impeded cell elongation. Binucleation.
characteristic of maturing cardiomyocytes, was reduced with soluble collagen supplement and
nanoscale topography and was associated with an increase in cytokinesis. Both nanoscale
topography and microcontact printing resulted in aligned cardiomyocyte monolayers but
produced different morphologies.
Lastly, the lessons learned from studying the aforementioned processes were applied to test the
utility of ESCDMs as biological actuators. Three proof-of-concept experiments were conducted:
ESCDM monolayers were able to contract synchronously as a cell-assemble, force generated by
the cell monolayer was estimated to be comparable to that by neonatal myocytes and lastly, the
direction of contraction could be controlled with surface patterning.
This work advances our understanding on the cardiogenic potential of murine embryonic stem
cells and elucidated complex biological questions with well-characterized and controlled tissue
engineering techniques.
Thesis advisor: Roger D. Kamm
Title: Singapore Research Professor of Biological and Mechanical Engineering
Acknowledgements
First and foremost, I would like to thank my advisor, Professor Roger Kamm. Working with
Roger was a bit like swimming in the ocean. I was given the freedom to roam around in the rich
and immense sea of knowledge and go at my own pace to figure out what I would like to do. At
the same time, when I was drowning in too many experiments that did not work, his
encouragements kept my afloat and his advice, like a lighthouse, always pointed me in the right
direction. His focused, yet laissez-faire advising style gave me a sense of ownership of the
project and taught me how to be a great researcher. But most importantly, Roger was a true
teacher who cared greatly about my personal growth as an individual. He let me take a summer
off to explore West Africa and gain some business skills. During my graduate career, I had the
opportunity to grow both scientifically and personally. For that, I cannot think of a better mentor
and advisor. There was also another perk to being Roger's student and that was the chance to
play with the sweetest and smartest dog I have ever met. I looked forward to every opportunity to
dog-sit Oakley and am pretty certain I always had more fun that he did. For keeping a smile on
my face, this thesis is also for Oakley.
I want to also thank my committee members, Dr. Richard T. Lee and Professor Alan
Grondzinsky, for their patience and advice and collaborators from the Pruitt lab at Stanford
University and Charles Stark Draper Laboratory. Thanks for enriching my intellectual repertoire
and providing me with research suggestions and tools to accomplish my project.
I would also like acknowledge all my friends, especially those who keep me sane, grounded and
inspired. Here are some of the people who amazingly put up with me over the years - Mark,
Nate, Anusuya, Yannis, Aida, Sid, Ryo, Jessie, Priam, and Levi. Thanks for being the support to
keep me going, the shoulder to cry on when I felt depressed, the dreamer to keep me inspired, the
thinker to envision the future, the critic when I needed a reality check, the joker to keep me
laughing, and, above all, thanks for being a great friend. You have made this journey pleasant,
stimulating and incredibly memorable.
I want to acknowledge all the volunteers at PATS (Pediatric AIDS Treatment Support),
particularly Eliza, Alisha, Kristin, Jialan, and Kali. I have the privilege to work with a group of
incredibly talented young women on a cause that is dear to my heart. They inspire me to work
harder, be more generous and be a better person.
Lastly, I would like to send my most sincere gratitude to my family for their unwavering support.
For my mom, now you can finally read my thesis and understand why I said you wouldn't
understand what I do. For my dad, the original Dr. Wan, your philosophy and advice have
shaped me to be who I am. For my dearest sister and mother of two of the cutest boys on earth,
thanks for believing in me and constantly offering your cat's cells for my experiments. Thanks to
all the extended family members, uncles, aunts, cousins, and grandmother, for always being there
for me.
The past five years have been transformational and full of blessings. It would not have been
nearly as enjoyable without all of you. I hope I can continue on with all the support and love and
make the world just a little bit better, one cell at a time. @
"When we become more fully aware that our success is due in large measure to the loyalty,
helpfulness, and encouragement we have received from others, our desire grows to pass on
similar gifts. Gratitude spurs us on to prove ourselves worthy of what others have done for us.
The spiritof gratitudeis a powerful energizer." - Wilferd A. Peterson.
Table of Contents
Abstract
2
Acknowledgements .......................................................................................................................
3
Table of Figures.............................................................................................................................
9
Table of Tables............................................................................................................................
11
Chapter 1. Cardiomyocytes and Cardiogenesis ................................................................
12
1.1.
Cardiomyocytes in the Native Myocardium ...............................................................
12
1.2.
Cardiovascular Disease and Therapeutic Options.......................................................
16
1.2.1.
Muscle Based Treatment Options ....................................................................
1.2.2.
Adult Stem Cell and Other Progenitor Cell Based Treatment Options............ 17
1.2.3.
Embryonic Stem Cell Based Treatment Options .............................................
16
18
1.3.
Utilization of the Contractile Function of Cardiomyocytes for Miniature Machines .... 19
1.4.
Induced Organization of Cardiomyocytes..................................................................
24
1.4.1.
Protein-based Patterning ..................................................................................
24
1.4.2.
Topographical Patterning ................................................................................
24
1.5.
Applications of Hydrogel Scaffolds on Cardiomyocytes...........................................
28
1.6 .
C onclu sion s ....................................................................................................................
30
Chapter 2.
Generation of Cardiomyocytes Derived from Murine Embryonic Stem Cells
and the Study of Differentiation in a Compliant Microfluidic Platform*.........................
31
2 .1.
In trodu ction ....................................................................................................................
31
2 .2 .
B ack ground ....................................................................................................................
32
2.2.1.
Use of Microfluidic Platform in in Vitro Cell Culture ....................................
32
2.2.2.
Methods to Generate Embryoid Bodies (EBs)................................................
33
2.2.3.
Effect of Mechanical Stretch on Cardiomyocytes in Vitro .............................
34
2.2.4.
Effect of Mechanical Forces on Embryonic Stem Cell Differentiated
C ardiom yocytes in Vitro.....................................................................................................
5
35
2.3.
Materials and Methods ...............................................................................................
36
2.3.1.
Maintenance of Murine Embryonic Stem Cells.............................................
36
2.3.2.
Induction of Differentiation and Maintenance of Endothelial Cells ................
36
2.3.3.
Embryoid Body Culture Condition ..................................................................
37
2.3.4.
Microfluidic Device Design.............................................................................
37
2.3.5.
Stretch Application Apparatus ........................................................................
39
2.3.6.
Other Stretch Application Considerations.......................................................
40
2.3.7.
Image Analysis, Quantification and Statistical Analysis .................................
41
2 .4 .
R esu lts ............................................................................................................................
2.4.1.
Validation of Motorized Microfluidic Platform..............................................
2.4.2.
Differentiated Embryoid Bodies Exhibit Distinctly Different Morphologies in
Microfluidic Environment vs. Conventional Well Plates..................................................
2.4.3.
41
41
43
Microfluidic Environment Enhances the Differentiation of Embryonic Stem
Cells into Cardiomyocytes..................................................................................................
44
46
2.4.4.
Effect of Bone Morphogenetic Protein 2 ........................................................
2.4.5.
Uniaxial Cyclic Stretch Inhibits the Differentiation of Embryonic Stem Cells into
C ardiom yocytes .....................................................................................................................
2.5.
Discussions and Conclusions ......................................................................................
Chapter 3.
47
49
Effect of Extracellular Matrix Protein and Surface Patterning on Driving
Maturing Phenotypes of Isolated Embryonic Stem Cell Derived Cardiomyocytes during
Extended Culture**....................................................................................................................
52
3.1.
In trodu ction ....................................................................................................................
52
3 .2 .
B ack ground ....................................................................................................................
53
3.2.1.
Collagen I in the Myocardium ........................................................................
53
3.2.2.
Effect of Surface Patterning on Cardiomyocytes and Cardiogenesis ..............
54
3.2.3.
Maturation of Embryonic Stem Cell Derived Cardiomyocytes ............
54
3.3.
Materials and Methods ................................................................................................
57
3.3.1.
Embryoid Body Dissociation and Fluorescent Activated Cell Sorting............ 57
3.3.2.
C ell Seeding ....................................................................................................
59
3.3.3.
Microcontact Printing (MCP) ........................................................................
60
3.3.4.
Nanopattern Topography (NPT) ......................................................................
61
3.3.5.
Immunofluorescent Staining ...........................................................................
61
3.3.6.
Quantification and Statistical Analysis of Binucleation, Cell Shape and Cell
Alignment 62
3.4 .
R esu lts ............................................................................................................................
3.4.1.
64
Sarcomere development of ESCDMs suspended in collagen gel or plated on
ECM-coated polystyrene ....................................................................................................
64
3.4.2.
Cell Adhesion with Overladen Collagen Gel Layer.........................................
65
3.4.3.
Effect of Collagen I on the Elongation of ESCDMs.......................................
66
3.4.4.
Effect of Collagen I on Binucleation................................................................
67
3.4.5.
Effect of Surface Patterning on ESCDM Alignment ...................
69
3.4.6.
Effect of Surface Patterning on ESCDM Elongation...................
72
3.4.7.
Effect of Surface Patterning on ESCDM Binucleation.................
73
3.5.
Discussion and Conclusions.......................................................................................
75
Chapter 4. Synchronization, Force Generation, and Control of Contractile Direction of
Embryonic Stem Cell Derived Cardiomyocytes.......................................................................
79
4 .1.
Introdu ction ....................................................................................................................
79
4 .2 .
B ack ground ....................................................................................................................
80
4.2.1.
Utilizations of Cardiomyocytes as Biomachines in 2D ...................................
80
4.2.2.
Utilizations of Cardiomyocytes as Biomachines in 3D ...................................
81
4.3.
Materials and Methods ................................................................................................
4.3.1.
Cell Culture and Isolation ...............................................................................
7
82
82
4.3.2.
In-phase Contraction Analysis ........................................................................
82
4.3.3.
Force Generation .............................................................................................
83
4.3.4.
Control over Contractile Directions ...............................................................
84
4 .4 .
R esu lts ............................................................................................................................
4.4.1.
In-Phase Contraction is Detected across Monolayer.......................................
4.4.2.
Force Generated by Embryonic Stem Cell Derived Cardiomyocytes are
Comparable to That of Neonatal Myocytes.......................................................................
4.4.3.
4.5.
Direction of Contraction can be Controlled with Pattern Directions ...............
Discussions and Conclusions ......................................................................................
Chapter 5.
Conclusions............................................................................................................
5.1.
Significant Findings and Summary .............................................................................
5.2.
A re We There Yet? ........................................................
5.3.
Future D irections..........................................................................................................
............................................ ..
86
86
87
90
93
97
97
10 1
105
5.3.1.
Microfluidic Platform Development .................................................................
105
5.3.2.
Further Investigation into Binucleation.............................................................
105
5.3.3.
Incorporation of Additional Stimulations .........................................................
106
5.3.4.
Investigation of Maturation of ESCDM Electrical Properties ............
107
5.3.5.
Incorporation of Patterning in 3D and Other Cell Types ..................................
108
5.3.6.
Self A ssembly ...................................................................................................
108
5.3.7.
Concluding Thoughts ........................................................................................
109
Appendix I - Embryonic Stem Cell Maintenance and Differentiation Protocol........ 110
Appendix II - Cardiomyocyte Enrichment with Percoll Separation...................................
References
114
113
Table of Figures
Figure 1-1: Cardiomyocyte Orientationin Native Myocardium..............................................
13
Figure1-2: DecellularizedHeart..............................................................................................
14
Figure 1-3: Microdevices Powered by Cardiomyocytes ...........................................................
23
Figure 1-4: Alignment of Neonatal Myocytes with MicrocontactPrinting and Nanoscale
Topography ...................................................................................................................................
26
Figure 1-5: ProposedMechanism of Nanotopography...........................................................
27
Figure 2-1: M icrofluidic Design................................................................................................
38
Figure 2-2: Calibrationof PDMSfilm thickness as a function of spin coater rotationalspeed.. 39
Figure 2-3: Uniaxial Stretch Platform ......................................................................................
40
Figure 2-4: Validations of the ProperFunctions of the Microfluidic Device and the Cyclic
Stretch Pla tfo rm ............................................................................................................................
43
Figure2-5: MorphologicalDifferences of Embryoid Bodies..................................................
44
Figure2-6: Method of Differentiationand Representative Images of Differentiated
Cardiomyocytes.............................................................................................................................
45
Figure2-7: Microfluidic Environment Enhances Differentiation...........................................
46
Figure2-8: Effect of BMP2 on Differentiation.........................................................................
47
Figure2-9: Uniaxial Cyclic Stretch Inhibits Differentiation....................................................
48
Figure3-1: Cardiomyocytes at Various Developmental Stages................................................
55
Figure3-2: FluorescentSorting Results...................................................................................
58
Figure3-3: Schematics of Cell Seeding Conditions..................................................................
60
Figure3-4: Cell Boundary Identification and Verification with N-Cadherin..........................
62
Figure3-5: ESCDMs Embedded in Collagen I Hydrogel.........................................................
64
Figure3-6: Representative Images of ESCDMs on Polystyrene..............................................
65
Figure 3-7: Cell Adhesion to Overlaying Collagen Gel...........................................................
66
Figure 3-8: Effect of Collagen I on Aspect Ratio ......................................................................
67
Figure 3-9: Effect of Collagen I on Binucleation......................................................................
68
Figure 3-10: Effect of Collagen I on Binucleation on PatternedSurfaces................................
69
Figure 3-11: Cell Alignment with Microcontact Printingand Nanopattern Topography........ 70
Figure 3-12: Histogram of Cell Alignments .............................................................................
71
Figure3-13: Effect of Surface Patterningon Elongationwithout the Addition of Collagen....... 72
9
Figure 3-14: Elongation of ESCDMs with Soluble Collagen and Collagen Gel ......................
73
Figure 3-15: Effect of Surface Patterningon Binucleation.......................................................
74
Figure 4-1: ParticleTracking to Investigate Synchronization..................................................
83
Figure 4-2: FluorescentBeads to Estimate Force Generation...............................................
84
Figure 4-3: Schematic Illustration of the Effect of Anisotropy on Contraction Direction........ 85
Figure 4-4: In Phase Contraction of Cell Monolayer .............................................................
87
Figure 4-5: Movement of a FluorescentBead...........................................................................
88
Figure 4-6: Schematic of Force Generation..............................................................................
90
Figure 4-7: ESCDMs Contractalong the Directionof Surface Patterning.............................
91
Figure4-8: Different Methods of Measuring Forces of Cardiomyocytes.................................
95
Figure 5-1: Sarcomere organizationof ESCDMs compared with adult cardiomyocytes.......... 102
Figure 5-2: Time course of phenotypic changes of ESCDMs.....................................................
104
Table of Tables
Table 1: Bioactuatorspowered by aggregatesof cardiomyocytes in 2D or 3D ......................
21
Table 2: Force measurements of cardiomyocytes ....................................................................
95
Table 3: Effect of MechanicalStretch on Differentiation...........................................................
107
Chapter 1. Cardiomyocytes and Cardiogenesis
Research on cardiomyocytes has numerous important implications. Cardiovascular diseases are
the leading cause of mortality in the world (1) and many fundamental mechanisms of such
diseases remain elusive. From a basic science perspective, the highly-ordered hierarchy, from
myosin-actin pair, sarcomere organization, myotube formations to the large myocyte sheet
structures, calls attention to the fascinating underlying engineering principles in biology. In this
chapter, we provide the background necessary to motivate this thesis work. It is structured to
present information ranging from a biological standpoint to a more engineering application
perspective, from understanding the structure of the native myocardium to in vitro robotic
machines powered by cardiomyocytes.
1.1.
Cardiomyocytes in the Native Myocardium
The heart is a chemical-mechanical-electrical biological organ. To ensure the accurate electrical
conduction and subsequent contraction, Precise and intricate cell organization has to be
maintained. By volume, cardiomyocytes make up 70% of the cell volume in the heart (2). The
rest of the cells, such as endothelial cells and cardiac fibroblasts, are critical in adequate
nutrient/waste transport and maintenance of the structural integrity by producing adequate
extracellular matrix proteins (3,4). From tissue dissections and SEM images, it can be seen that
cardiomyocytes (- 4 myocytes thick) are layered as sheets with specific orientations to produce
the observed coordinated contractions (5). Helm et. al. used Diffusion Tensor Magnetic
Resonance Imaging to illustrate the fiber orientations of a canine heart and found similar
orientation results (Figure 1-1) (6). All these studies support the notion that aligned
cardiomyocytes with specific orientations are essential in producing the precise twisting motion
during each contraction (7,8).
.............
(b)
(a)
V11
(c)
Figure 1-1: Cardiomyocyte Orientation in Native Myocardium
(a) and (b) are reproduced from Legrice et. al. (5). With SEM, they demonstrated high anisotropy in
the left ventricle of a canine heart. Cardiomyocytes are arranged in sheets with 1-2 myocyte gaps in
between (Scale bar: 100pm). (c) Helm et. al obtained similar results for fiber orientation with
Diffusion Tensor Magnetic Resonance Imaging (6).
In addition to cells, the extracellular matrix (ECM) is equally important. ECM not only acts as a
supporting scaffold and a microenvironment for cell-cell interactions, but it also signals to the
cells via integrins and focal adhesion complexes (9). The composition of the extracellular matrix
is dynamic both temporally and spatially (10,11). Although the exact composition remains
unknown, it has been shown that the collagen content increases over time and the activation of
aisi integrin, which primarily binds to collagen I, is necessary for forming the rod-shape
phenotype of cardiomyocytes (12). Spatially, the composition and specific orientation of the
13
...........
-
collagen matrix play a significant role in maintaining the orientation of the cells and
accommodating for the mechanical strains during each contraction (4,13). Furthermore, the
collagen fibril network is important in preventing cardiomyocytes from over stretching (14,15).
Ott et. al. decellularized mouse hearts with an extensive SDS washing procedure preserving the
intricate ECM network comprised of collagen I, collagen III, laminin and fibronectin (Figure 1-2)
(16). The shape of the decellularized heart remains intact suggesting the ECM network, rather
than cellular constituents, is responsible for the structural integrity. Neonatal myocytes and
endothelial cells were introduced into the decellularized tissues and the recellularized hearts were
cultured with pulsatile medium perfusion and electrical stimulation mimicking the physiological
condition. Even so, the left ventricle systolic pressure was less than 25% of a fetal heart and 2%
of an adult heart (17). The limited pumping capability highlights the need for further
optimizations of cell seeding protocol and culture conditions.
Figure 1-2: Decellularized Heart
A cadaveric heart was decellularized with SDS for 12 hours. The remaining network consisted of
mostly collagen fibrils (16).
In summary, the heart is comprised of various cell types and extracellular matrix proteins.
Cardiomyocytes, the cells responsible for contraction, are highly organized, layered as 2D sheets,
14
and intimately surrounded by extracellular matrix, which is highly dynamic spatially and
temporally. All components are essential in maintaining the proper function of the heart. CVDs
usually lead to the maladaptations of the extracellular matrix and loss of cellular functions.
1.2.
Cardiovascular Disease and Therapeutic Options
To step back and examine the prevalence and significant disease burden of CVDs, one easily
comes to the conclusion that better and more widely available treatments are urgently needed. A
large portion of CVDs is attributed to myocardial infarction (MI) from coronary artery
occlusions. MI results in rapid and massive cardiomyocyte death. On average, 1 billion
cardiomyocytes undergo cell death in an MI (18). Effective replenishment of functional
cardiomyocytes along with revascularization is the Holy Grail of numerous cell therapy
strategies.
Current clinical treatments after a severe MI include surgical procedures, such as implantation of
left ventricle assisting device (LVAD) or a heart transplant. Neither of them is ideal and both are
very costly. As CVDs become a global issue, a major paradigm shift is necessary to repair the
damaged tissue and restore its function. Cell-based therapy is a novel approach and has shown
initial promises (19-23). In the following section, I will briefly summarize the current state-ofthe-art on different cell-based therapies.
1.2.1.
Muscle Based Treatment Options
Unlike post-mitotic cardiomyocytes, skeletal muscle cells and their progenitor cells proliferate.
Myoblasts respond to injuries and differentiate into skeletal muscle cells. Researchers have
attempted to replenish an infracted area with skeletal muscles or precursor cells with inconsistent
successes. Taylor et. al. reported a beneficial effect of skeletal myoblast transplantation (24)
while Menasche et. al. failed to exhibit statistically significant effects in a double-blind six
months study (25). Although the mechanism of the transient benefits of myoblast transplantation
remains unanswered, several challenges are more consistently reported. Massive cell death is
reported within the first hour of implantation, and surviving muscle cells do not express cardiac
16
specific proteins (26). Skeletal muscle and myoblast treatments are further complicated with
induced arrthymias and tachycardia (27). While the possibility of autologous transplantation
makes this treatment option attractive, proper cell homing, integration, trans-differentiation and
survival have proven to be challenging.
1.2.2.
Adult Stem Cell and Other Progenitor Cell Based Treatment Options
Most adult stem cells are multipotent, able to self-renew and differentiate into several, but not all,
cell lineages. Until recently, cardiogenesis with adult stem cells has focused on the potential
transdifferentiation of mesenchymal stem cells (28,29). Cardiac stem cells and induced
pluripotent stem cells are discovered within this past decade offering new treatment possibilities.
Adult progenitor cells such as bone marrow derived stem cells, mesenchymal stem cells (MSCs),
and adipose stem cells have been studied for their potential in differentiating
into
cardiomyocytes. MSCs naturally differentiate into osteoblasts, chondroblasts and adipocytes
when exposed to the appropriate stimuli. There have been some reports of transdifferentiation
(30). However, it is unclear whether transdifferentiation actually occurred or the results were
obscured by cell fusion events (31-33). While there is no consensus on MSC's capacity to
transdifferentiate, the transient benefits of MSCs have been largely speculated to originate from
the paracrine effects in increasing capillary density around the infarct zone or tissue mass to
reduce the stress on the failing myocardial wall (34-36). Regardless of the exact mechanism,
repeatable and long term improvements of cardiac function with MSC implantation remain
elusive (37,38).
Since transdifferentiation is controversial, much attention has been paid to progenitor cells with
the known capacity to differentiate into cardiomyocytes. Cardiac stem cells have been
discovered in the past decade and there has been a lot of interest in understanding and isolating
cardiac stem cells for regeneration (39,40). In animal studies, there was wide variance on how
these stem cells assisted in restoring cardiac function (41-44). The variability came largely from
the differences in cell isolation and identification techniques. More importantly, it also came
from the limited understanding of mechanisms of these cells in the myocardium. Much more
fundamental understanding is required before cardiac stem cells can be fully utilized as a
treatment option.
1.2.3.
Embryonic Stem Cell Based Treatment Options
Cell therapy with embryonic stem cell derived cardiomyocytes (ESCDMs) has been investigated
because their ability to self-renew and differentiate into all cell types. Unlike skeletal muscle
precursors, ESCs have been successfully differentiated into cardiomyocytes with similar
phenotypes and characteristics as fetal cardiomyocytes (45,46). ESC injection into the
myocardium induces the formation of teratomas (18).
Laflamme et. al. demonstrated
electromechanical coupling of transplanted ESCDMs into the host myocardium at 4 weeks but a
separate study showed no benefit of transplantation at 12 weeks (47,48). The conflicting and
complex findings on ESCDMs illuminate the need for a fundamental mechanistic understanding
and characterization of the cardiogenesis process of embryonic stem cells (49).
1.3.
Utilization of the Contractile Function of Cardiomyocytes for Miniature
Machines
In the previous sections, in vivo cellular and extracellular compositions were described. The
therapeutic importance of stem cell derived cardiomyocytes as a new source of functional
cardiomyocytes after injuries was also discussed. In this section, the focus is shifted and the
potential to utilize cardiomyocytes as a source of contractile units is explored.
Neonatal cardiomyocytes, harvested within 2 days after birth, are the gold standard for the cellpowered miniature machines because, unlike adult cardiomyocytes, neonatal myocytes contract
spontaneously and continuously and do not lose their phenotype in vitro. Tanaka et. al. seeded
neonatal cardiomyocytes on the surface of a hollow PDMS sphere and on a thin PDMS sheet to
devise spherical and linear pumps respectively (Figure 1-3 a&b) (50,51). For the hollow PDMS
sphere, the contractile force generated by the myocyte sheets was strong enough to squeeze the
sphere creating a pulsatile flow through the glass capillaries connected to the sphere. This device
was extremely delicate and difficult to produce and while the objective was to create a heart-like
flow chamber, it resembled little of the heart. Kim et. al. similarly seeded neonatal
cardiomyocytes on 2D to construct a walker (Figure 1-3c) (52). The walker was fabricated with
soft-lithography techniques and capable of propelling forward for over one week. This hybrid
design, albeit sophisticated in fabrication, lacks real world applications. Feinberg et. al. induced
the alignment of neonatal myocytes on a thin PDMS film microcontact printed with 20/2Opm
line patterns of high (50ig/ml) and low (2.5pg/ml) concentrations of fibronectin. By cutting the
monolayer sheet along the direction of cell alignment, the monolayer produced a forward motion
and in contrast, when the rectangular monolayer sheet was cut at a 5-15' angle from the cell
alignment, it resulted in a twisting motion (Figure 1-3d) (53). This study is one illustration in
19
taking advantage of the intrinsic cell responses to extracellular matrix protein variations to
produce efficient bioactuators.
In addition to utilizing cardiomyocytes by seeding them on a 2D substrate, Xi et. al. combined
material science to seed cells in innovative arrangements (Figure 1-3 e&f) (54). They used the
thermoresponsive polymer, poly N-isopropylacylamide (PNIPAAM), with the Lower Critical
Solution Temperature (LCST) at 32'C, below which PNIPAAM is hydrophilic. In room
temperature, microstructures, coated with gold, were embedded in PNIPAAM hydrogel. Cells
were subsequently seeded on the hydrogel with a preferred attachment to the gold surface. Upon
cell adherence, the system was heated above LCST resulting in PNIPAAM becoming
hydrophobic and the hydrogel collapsing leaving cells attached only to the microstructures.
Last category of machines mixed neonatal myocytes or ESCDMs in an extracellular matrix
hydrogel (Figure 1-3 g&h) (55,56). Cell-gel mixture was polymerized in a toroid-shaped mold.
Once it was polymerized, it was released and stretched to induce alignment. After an extended
period of stretch (>7 days), the cell-gel band was able to spontaneously contract, albeit
asynchronously.
In this section, we provided a summary of existing designs of micromachines powered by
cardiomyocytes - either by seeding them in 2D, utilizing unique thermo-responsive materials or
mixing with 3D hydrogel. The limitations and force output of each design are listed in Table 1.
Specifically, many 2D experiments were terminated within 1 week and phenotypic changes were
observed without further investigation (50). The pumps and the walker did not utilize the
intrinsic behaviors of cardiomyocytes, such as cell alignment or synchronization. Instead, it
deployed sophisticated fabrication techniques and used cardiomyocytes only as a miniature
actuator. This approach generally produces much lower force. In Chapter 4, we will explore how
to harness the understanding of cardiomyocytes as a cell population into implementing
microdevices.
70 pN*
11 days
Difficult to fabricate
Require PDMS support
Lack of feasible therapeutic
applications
Small forces
Unaligned cell monolayer
Noatal
60 pN
Upuire
2KPa
3 days
Q.51mN (Fmax)
>26 days
PDMS support
Lack of feasible therapeutic
applications
Short term experiments
Dif~uttg44Myceapurity
Myopytes
Toroid~oskalwts Mysologil
(55)
relevance
Decellularized
and recellularized
heart
Neonatal
myocytes
320Pa
8 days
Adult mnyocytt~s
l 4Ka**
Adult'
Small pressure
(16)
Wtvetr~e
vivo
350mL/min per Fetal
Fetal ventricle(57) Fetal myocytes
lamb in
kg
vivo
Table 1: Bioactuators powered by aggregates of cardiomyocytes in 2D or 3D
*: the forces are estimated as Stokes' drag force assuming the kinetic viscosity of medium to be
0.7x10-3 NsIm 2; **: determined by estimating forces required for PDMS deflection ***: assuming
LVEDP-10mmHg and peak stress -120mmHg
Uh
N,)
()
F)
Figure 1-3: Microdevices Powered by Cardiomyocytes
(a,b) a spherical pump and a vertical pump produced by Tanaka et. al. (50,51) (c) PDMS-based
walker developed Kim et. al. (52) (d) patterned neonatal myocyte sheets by Feinberg et. al. (53). (e,f)
A myocyte cantilever and a walker developed by Xi. et. al. with PNIPAAm and microfabrication (54).
(g) three-dimensional engineered heart tissue (EHT) (55). (h) Guo et. al. developed similar 3D
hydrogel-cell construct (56).
1.4. Induced Organization of Cardiomyocytes
One pertinent tissue engineering technique to control cell alignment is surface patterning.
Cardiomyocytes and the underlying extracellular matrix network are both highly anisotropic in
vivo and to reproduce similar anisotropy in vitro, two techniques have been widely used patterning with varying protein distributions or imposing direct topographical cues.
1.4.1.
Protein-based Patterning
Microcontact printing technique was developed in the Whitesides and Ingber laboratories in the
mid-90's (58) and since then, it has been widely used to control cell shapes and functions. Line
patterns of alternating concentrations of ECM proteins are often used to produce aligned
monolayers, as Feinberg et. al. did with fibronectin (Figure 1-4 a&b) (53). Variations in line
widths (12.5pm/12.5pm lines vs. 25pm/25pm lines) have been tested and no statistically
significant differences in the electrophysiological properties of neonatal myocytes was found by
measuring the conduction velocities, wavelengths and action potential durations in the
longitudinal and traverse directions (59). This study suggests that the exact dimensions of protein
patterning, within a reasonable range, do not affect cell behaviors. Microcontact printing, unlike
topographical cues, does not induce additional geometric constraints on cells which can
complicate analyses. However, since many cells secrete their own extracellular matrix proteins,
the efficacy of microcontact printing is likely to be transient.
1.4.2.
Topographical Patterning
With the advancement of microfabrication techniques, nanometer-scale patterns can be made and
this has been used for biological applications. Cells seeded on top of the nanotopography align
along the direction of the patterns (Figure 1-4 c&d) (60). The exact mechanism of the alignment
is still unclear but one speculation assumes the alignment is a result of integrins distributing
24
along the ridges. Kim et. al. recently proposed a different hypothesis suggesting that the effect of
nanoridges and grooves is attributed to an increase of cell spread area and contraction-mediated
stress (Figure 1-5a) (61). Cell spread area of neonatal myocytes were significantly larger on
800nm/800nm ridges/grooves surface compared with 400nm/400nm surface or unpatterned
surface because cells seeded on 800nm/800nm surface extended and spread into the grooves
while those on 400nm/400nm surface did not (Figure 1-5b&c). Despite the phenotypic
differences, the conduction velocities of the two conditions were not statistically different.
Furthermore, whether or not nanoscale topographical cues exist in vivo remains controversial.
Nevertheless, nanoscale topography has the advantage over microcontact printing in providing
the alignment cues over a long period of time.
Cell sensitivities to variations in nanotopography (periodicity, duty cycle and depth of the
grooves) are unknown. Au et. al. demonstrated qualitatively clearer sarcomeres for neonatal
myocytes seeded on a 1pm ridge/lpm groove surface compared with that on a 1pm ridge/4pm
groove surface although the cause for this difference remains elusive (60). Additional studies are
necessary to quantify the effects of nanotopography on cells.
Figure 1-4: Alignment of Neonatal Myocytes with Microcontact Printing and Nanoscale Topography
(a,b) Cardiomyocytes aligned along 20pim/20pm line pattern of alternating concentrations of
fibronectin (53). (c,d) Nanoscale topography was imposed on neonatal myocytes to induce alignment.
The grooves and ridges were 1pm wide and 400nm high. (60).
(a)
tColl penetration
I nto dhe grooves
Cefi-subsatum
adhesion
17,9W,
Cell
spreading
I i I
Fa.m
Con&acuon
mediatedsmas
I
TCx43
WCiil coupling
I Conduction velocity
(b)
(c)
Figure 1-5: Proposed Mechanism of Nanotopography
Kim et. al. suggested that cellular response was greatly impacted by the nanoscale topography (61).
(a) a schematic of their hypothesis was illustrated. Areas of direct plasma membrane-substratum
contact are highlighted in green and gap junctions are shown in red. (b,c) SEM image of neonatal
myocytes seeded on 800nm/800nm and 400/400nm ridge/groove patterns respectively. Scale bar:
200nm
1.5.
Applications of Hydrogel Scaffolds on Cardiomyocytes
In addition to surface patterning, another commonly-used tissue engineering technique is the
incorporation of hydrogel scaffolds. The hydrogel can be naturally-present ECM components
such as collagen I or basement membranes, or synthetic materials, such as polyacrylamide or
self-assembly peptide gels. Several groups have used hydrogels to study cardiomyocyte
properties in vitro. One common application is to use hydrogels as a cell culture supporting
scaffold. Leung et. al. used a mixture of collagen I and Matrigel TM to study endothelial cellneonatal myocyte co-culture and found the cell sheet to be electrically responsive (62). In our
laboratory, we have also demonstrated that co-culture of endothelial cells and neonatal myocytes
on a synthetic peptide gel promoted cardiomyocyte survival (63).
Another property of in vitro hydrogel scaffolds is the tunability of the substrate stiffness. Bhana
et. al. cultured neonatal myocytes on polyacrylamide gels of different stiffnesses by varying
cross-linking densities (64). They found that cardiomyocytes displayed the clearest striations and
the highest contractile force development when cultured on a substrate with similar stiffness as
the native myocardium.
In addition to providing the structural support, the composition of ECM hydrogel is critical to the
activations of integrins for cell binding and further mechanotransduction responses. For example,
EHTs developed by Zimmermann et. al. is comprised of a mixture of collagen I, Matrigelm, and
chick serum which activates multiple integrins (55). The engagement and activation of multiple
integrins, such as aip,, a 3Pi, a15 i and a7 pi, are necessary for the formation of a coherent
cardiomyocyte structure as neonatal myocytes and ESCDMs suspended in collagen I alone are
unable to extend and form cell-cell contacts (65).
These are several examples of how in vitro tissue engineering tools can be used to characterize
cell behavior. In Chapter 3, we will investigate the effect of collagen hydrogel on the phenotypic
changes of embryonic stem cell derived cardiomyocytes.
1.6.
Conclusions
This chapter provides the necessary background to motivate this thesis work whose overarching
aim is to obtain a better understanding of the cardiogenesis process in vitro. This is essential
because of the increasing epidemic of cardiovascular diseases and the lack of viable and
affordable options. Stem cell based therapies have been studied for over 25 years and no
consistent success has been reported. Most in vivo studies are plagued with high complexities
(e.g. cell survival, integration and safety) and experimental artifacts (e.g. cell fusions) (18,49). In
vitro characterizations are a powerful tool to systematically elucidate individual stimuli on the
cardiogenesis process.
To our knowledge, it has not been possible to produce mature cardiomyocytes from
undifferentiated embryonic stem cells in vitro. Several key questions still remain to be answered:
1. how is differentiation affected by in vitro culture conditions? and 2. how do nascently
differentiated cardiomyocytes develop more mature phenotypes? In the following two chapters,
we separately addressed the two questions. First we examined how the differentiation process
could be modulated with the changes in culture dimension and mechanical stimulations. Second,
we extended the investigation to characterize the morphological changes of differentiated and
isolated cardiomyocytes with the presence of collagen I and surface patterning and showed that
these biophysical factors can influence the maturing phenotypes. Lastly, we explored the
possibility to use these cells for non-medical applications. The findings of this thesis have direct
implications on how differentiation and maturation can be augmented with the appropriate
biophysical stimuli.
Chapter 2. Generation of Cardiomyocytes Derived from Murine Embryonic
Stem Cells and the Study of Differentiation in a Compliant Microfluidic
Platform*
*
The majority of the work presented in the chapter has been submitted.
2.1.
Introduction
In this chapter, the differentiation of embryonic stem cells into cardiomyocytes, a process termed
cardiogenesis, was investigated. Studying cardiogenesis is important since embryonic stem cells
have been considered as a therapeutic means for cardiovascular diseases. As described in the last
chapter, there is limited understanding of how cardiogenesis is affected by biophysical
stimulations. Here, cardiogenesis was investigated with a compliant microfluidic platform which
allows for versatile cell seeding arrangements, optical observation access, long term cell viability,
and programmable uniaxial cyclic stretch. Specifically, two environmental cues were examined
with this platform - culture dimensions and uniaxial cyclic stretch. First, the differentiation
process was enhanced in microfluidic devices compared with conventional well-plates, partially
from the accumulation of cardiogenic factor, BMP2, in a confined space. Second, uniaxial cyclic
stretch at 1Hz was found to have a negative impact on differentiation. This microfluidic platform
builds upon an existing design and extends its capability to test cellular responses to mechanical
strain. It provides capabilities not found in other systems for studying differentiation, such as
seeding embryoid bodies in 2D or 3D in combination with cyclic strain. This study demonstrates
that the restricted transport in a microfluidic system contributes to enhanced differentiation and
may be a superior platform compared with conventional well plates. In addition to studying the
effect of cyclic stretch on differentiation, this compliant platform can also be applied to
investigate other biological mechanisms. The results in this study illuminate the importance of
31
biophysical cues on the differentiation process and provide guidance as to how these cues can be
effectively used to improve the differentiation process.
2.2.
Background
2.2.1.
Use of Microfluidic Platform in in Vitro Cell Culture
Microfluidic devices (ptFDs) are excellent in vitro systems in which to study cell functions, build
disease/organ models, and dissect mechanisms of specific stimulations in a systematic manner
(66). Previous work from our laboratory has demonstrated that pFDs can be used to examine
interactions of multiple cell types and effects of chemotaxis on angiogenesis and cancer cell
migration (67,68). The versatile design allows for cell seeding arrangements in both 2D and 3D,
application of shear stress or interstitial flow, and microscope access for continuous observation.
In this study, a modified device, capable of imposing periodic uniaxial stretch without sacrificing
the imaging capabilities, was developed to study the differentiation of embryonic stem cells
(ESCs) into cardiomyocytes. With this platform, we were able to study how cyclic stretch affects
the cardiogenesis process in a well-controlled microfluidic system.
Previous work has shown that murine cardiogenesis, involving the generation and manipulation
of embryoid bodies (EBs), can be augmented both biochemically and biophysically.
Biochemically, ascorbic acid, DMSO, retinoic acid, FGF and BMP2/4 are some of the growth
factors that have been demonstrated to promote cardiogenesis and new cardiogenic chemicals are
continuously being discovered (69-77). Biophysically, control of EB size, electromagnetic
stimulation and mechanical strain have also been shown to enhance cardiac differentiation (7881). In this study, we attempt to utilize a microfluidic system to impose biochemical and
biophysical stimulations to EBs.
Microfluidic platforms have been shown to affect diffusion-dominated processes (82). Yu et. al.
demonstrated that cell proliferation rate was dependent on the height of the microchannels,
presumably due to an accumulation of secreted factors. By comparing microchannels to
conventional cell culture well plates, higher proliferation rates have been observed for murine
mammary gland cells and during murine embryo development (83,84). Existing literature
suggests that the diffusion of growth factors enhances cell processes, such as proliferation, and
augments cardiogenesis, so we hypothesize that differentiation will be enhanced in the confined
space of microfluidic devices.
2.2.2.
Methods to Generate Embryoid Bodies (EBs)
EBs are aggregates of embryonic stem cells and are regarded as the gold standard of murine
embryonic stem cell differentiation (85). EBs are formed in a low stress suspension culture.
Different methods have been used to efficiently generate EBs of uniform sizes including hanging
drops, seeding within microwells, rotating bioreactors and culturing within non-adherent
scaffolds. These methods are superior to monolayer culture or single cell suspension culture with
higher EB yields and more homogenous EB sizes. Rotating bioreactors offer the possibility of
large-scale production (86) but require additional equipments. Hanging drops method can be
performed in any laboratory with cell culture facilities. Microwells and non-adherent scaffold
methods are ideal for designing EBs with specific sizes (78).
EBs develop three germ layers over time as observed in development. The differentiation lineage
can be directed by exposing EBs to the appropriate set of stimuli. While EBs are necessary in the
induction of differentiation, they pose unique challenges in transport and dissociation for further
characterizations (87). Considerations regarding to various dissociation techniques are discussed
in Section 3.3.1.
2.2.3.
Effect of Mechanical Stretch on Cardiomyocytes in Vitro
The effect of mechanical stretch on differentiated myocytes in vitro has been studied in great
detail. It has been demonstrated that cyclic stretch is required for neonatal myocytes suspended
in hydrogel to form a coherent tissue construct and develop sarcomeres and cell-cell contacts
(22). In vitro studies have also provided invaluable knowledge on how mechanical stress induces
protein expressions and/or activations. Bullard et al. found that protein kinase C (PKC)
phosphorylated differently depending on the direction of stretch, cyclic or static and the
magnitude of stretch (88). Furthermore, Mansour et. al. suggested that PKCc and focal adhesion
kinase (FAK) are necessary to restore the resting length of sarcomeres after uniaxial static strain
(89). Mechanical stress has also been demonstrated to affect the production and distribution of
junctional proteins such as connexin 43 and cadherins (90,91).
Mechanical stretch has also been proven to cause morphological changes in myocytes. One in
vitro study suggested that neonatal rat ventricular myocytes display adult myocyte characteristics
after two weeks of culture under mechanical stretch with an increased and better-organized
sarcomere structure (92). Another study found in vitro culture of cardiac myocytes align along
the direction of stress (88), a process that is apparently mediated by N-cadherin and Rac-1
(91,93). Yet, another study by Kada et. al. found that orientation of myocytes changes depending
on the time course of stretch stimulation (94). All current data on protein expression/activation
and cell morphology have unequivocally demonstrated the importance of mechanical stimulation
on differentiated cardiac myocytes.
2.2.4.
Effect
of Mechanical
Forces
on
Embryonic
Stem
Cell Differentiated
Cardiomyocytes in Vitro
While the mechanotransduction pathways of differentiated cardiomyocytes
have been
extensively studied, the effect of mechanical stimulation on the differentiation of embryonic
stem cells (ESCs) into cardiomyocytes and the mechanisms of action are less clear (95-97).
Schmelter et. al. suggest that mechanical stretch activates the reactive oxygen species signaling
pathway and thus enhances
the differentiation of murine embryonic stem cells into
cardiomyocytes (79). Opposite results indicating that stretch inhibits differentiation have also
been shown, attributed to the activation of TGF-p/Activin/Nodal pathway (98). These conflicting
results illustrate the need for further studies.
It is also important to note that current studies on EB cardiogenesis with stretch are limited to a
two dimensional seeding condition. Three-dimensional environments, however, resemble more
closely the native myocardial environment during development and myocardial infarct zones
targeted for stem cell therapy (99-101). Therefore, a microfluidic system which allows EBs to be
seeded in 3D and experience cyclic uniaxial stretch might provide valuable new insights into the
differentiation of embryonic stem cells into cardiomyocytes, and might also elucidate other
important cellular behaviors where mechanotransduction is implicated.
2.3.
Materials and Methods
2.3.1.
Maintenance of Murine Embryonic Stem Cells
Murine embryonic stem cells (mESC) expressing a cardiac specific a-MHC promoter that was
tagged with green fluorescent protein (GFP) (line CGR8, kindly provided by RT Lee, Harvard
Medical School) allowed direct observation of differentiation into cardiomyocytes. To maintain
ESCs in an undifferentiated state, Glasgow Minimum Essential Medium (GMEM) (Invitrogen),
supplemented with 1,000U/ml leukemia inhibitory factor (LIF, Sigma), 1mM Sodium Pyruvate
(Invitrogen), 1x Non-Essential Amino Acid (Invitrogen), 15% Knockout Serum Replacement
(Invitrogen), 25mM of HEPES,
10- 4M p-mercaptoethanol
(Sigma), and 1x Penicillin-
Streptomycin (Invitrogen) was used. Cells were maintained in flasks coated with 0.1% gelatin in
PBS. Cell confluency was tightly controlled not to exceed 70%. The detailed protocol can be
found in Appendix I.
2.3.2.
Induction of Differentiation and Maintenance of Endothelial Cells
By removing LIF and creating a three-dimensional environment, mESCs spontaneously
differentiated.
The composition of the differentiation medium was identical to that of the
maintaining medium except for the removal of LIF, the replacement of knockout serum by ESC
Fetal Bovine Serum (Invitrogen), and the addition of 100pM of ascorbic acid (74).
A standard hanging drop technique was used to induce differentiation (102). Briefly, cell
suspension solution was prepared at 10,000 cells/ml. 30pl drops were placed on the inside of a
100mm non-tissue culture treated Petri dish containing approximately 10ml of 1x PBS to prevent
evaporation. Drops, containing small cell aggregates, were cultured for 2 days before being
collected with a 10ml pipette. These aggregates were then cultured in differentiation medium for
3 more days for embryoid body (EB) formation.
Human microvascular endothelial cells (hMVECs, Lonza) were cultured with complete EBM-2
(Lonza). Passages 4-7 were used for experiments with stretch stimulation.
2.3.3.
Embryoid Body Culture Condition
Embryoid bodies (EBs) are aggregates of ESCs. For the 2D experiments, 5-8 EBs were seeded
into pFD channels or onto 12-well plates coated with 0.1% gelatin. To seed EBs in 3D, stock
collagen I solution derived from rat tail tendon (BD Biosciences) was mixed with 0.5N NaOH,
lOx DMEM, water, medium containing 500 EBs/ml to produce a pH 7.4, 2mg/ml collagen I gel
containing EBs. 10ptl of gel were used for each microfluidic device.
Specific device preparation protocol and gel filling techniques have been described previously
(67). Briefly, pFDs were permanently bonded with plasma and coated with 0.1% gelatin. Then
they were dried overnight in an 80'C oven to restore PDMS hydrophobicity. Channels were
filled with differentiation medium after collagen gel had fully polymerized.
2.3.4.
Microfluidic Device Design
Microfluidic devices (pFD) were used to study differentiation. pFDs have been designed in our
laboratory to study various biological processes, such as angiogenesis, cell migration, and liver
tissue engineering (67,68). The existing design consists of three fluid channels and two hydrogel
scaffold regions (Figure 2-la). It has the following key advantages:
1. Cells can be cultured either on 2D or in 3D,
2. Devices are made with polydimethylsiloxane (PDMS), a cell-benign and transparent
elastomer, and glass coverslips - allowing for cell live imaging, and
37
:::
....
.................
..
....
.....
.....
, , ..
3. Devices allow for adequate nutrient, gas, and waste transport.
Modifications were required in order to apply uniaxial cyclic stretch without sacrificing the
advantages of the existing design. These modifications included (Figure 2- lb&c),
1. Channel patterns were made on a 0.5-1mm thick PDMS layer, rather than 5-7mm,
2. Coverslips in the original design were replaced by a PDMS film of similar thickness, and
Rectangular, instead of circular, devices were used with additional clamp anchors.
(a)
(C)
Figure 2-1: Microfluidic Design
The schematics of the microfluidic device: the original design (a), the updated flexible device (b) and
the sample (c) of the final design which satisfied all the design criteria.
A spin coater was used to produce thin films of PDMS. The films had to be thin, pliable and at
the same time easy to handle with a tweezer. We calibrated the spin coater to determine the
relationship between rotational velocity and film thickness (Figure 2-2). In this study, PDMS
film was produced with the spin coater at 500 rpm for 30 seconds.
............
......
.................................................
PDMS Film Thickness
600
500 400 300 -
894
y = 56522x-0.
200 - R2 = 0.9726
100 0
,I
0
200
I
I
800
600
400
Rotational Velocity [rpm]
I
1000
1200
Figure 2-2: Calibration of PDMS film thickness as a function of spin coater rotational speed
PDMS with 10:1 base to curing agent was used to calibrate the thickness. PDMS was spun for 10
seconds to spread over the surface and 30 seconds at the indicated velocity.
2.3.5.
Stretch Application Apparatus
The actuator to apply cyclic stretch to the pFDs was chosen based on the following requirements:
" Max Travel Distance = (Maximal Strain, 20%) x (Total Length, 10mm) = 2mm
*
Accuracy = (Min Travel Distance) x (Allowable Error) = (Min Strain, 5%) x (Total
Length, 10mm) x (Allowable Error, 5%) = 12.5pm
" Max Velocity = Max Travel Distance x Max Frequency = 2mm x 10Hz = 20 mm/s
Parker MX80S precision linear motor (Irwin, PA) satisfied all the requirements. To prevent rust,
the motor was enclosed in a stainless steel box. Parker MX 80S is capable of withstand a thrust
load up to 123N, much higher than forces required to stretch 4 pFDs.
To connect pFDs to the linear motor, an in-house made clamp was used (Figure 2-3). The clamp
could accommodate up to 4 pFDs for one set of experiments. One side of the clamp was firmly
attached to the plates inside the incubator while the other side was connected to the linear motor.
I ..................
"I'll,
.
.....
........
(a)
Controller
Controlle
Stepper
Motor
Driver
Incubator (95% humidity,
37 0 C, 5% CO
-)
Linear
Precision
I
Stage
E
X
Commanded position vs. Time
E 1000
(b)
Anchor
a
Power
(b)
Microfluidic
Devices
5 0
0
-00
2
1
4
-1000
Figure 2-3: Uniaxial Stretch Platform
Design of the stretch platform: (a) schematic diagram illustrates that the stretch apparatus was
placed in the incubator; (b) displacement could be precisely controlled.
2.3.6.
Other Stretch Application Considerations
The most commonly used and commercially available stretch platform is the Flexcell* system
which imposes programmable equibiaxial cyclic or static strain. The system replaces the bottom
of a conventional 6-well plate with a flexible where cells can be cultured. This setup requires a
large amount of cells and also limits cell seeding to a 2D culture. Furthermore, while equibiaxial
strain is applicable in some physiologically relevant settings, it is more appropriate to apply
uniaxial stretch to mimic what cardiomnyocytes experience in the heart. Because of the limitations
of the need for a large number of cells, cell seeding conditions and stretch directions, we decided
40
to build an in-house system where a small amount of cells can be stretched uniaxially in a
microfluidic condition.
2.3.7.
Image Analysis, Quantification and Statistical Analysis
Both phase contrast and fluorescent images (20x) were taken with a Nikon Eclipse TE300
Microscope with Open Imagem Software. Individual embryoid bodies (EBs) were tracked and
observed daily for GFP expression with identical exposure settings. The first day that GFP
expression could be observed was defined as GFP1 and subsequent days as GFP2, GFP3 and so
on. Images were taken daily and analyzed with Matlab. Without any contrast enhancement, a
GFP-positive pixel was defined to be brighter than 120 on a 256 gray scale image. Most
differentiation occurred on the flat parts of the adherent EBs that have spread outwards so the
errors due to measuring the 2D projection were considered to be small. Still, direct comparisons
between the 2D and 3D experiments are subject to error.
All data are presented as mean ± SEM. The Student's t-test was used to identify statistical
significance (p<0.05). At least 20 EBs were examined and more than 3 independent samples
were used for each condition.
2.4.
Results
2.4.1.
Validation of Motorized Microfluidic Platform
The compliant tFDs were connected to a precision linear stage which could be programmed to
translate at specific frequencies and magnitudes (Figure 2-3). Cells seeded in the p1FDs were
stretched for 24 hours at 10% strain and 1Hz after they have fully adhered after 3 days.
Two different validation tests were initially conducted to ensure that cells actually experienced
the imposed mechanical stimulation. First, we confirmed that the gel could withstand cyclic
41
stretch without fracturing or detaching from the PDMS walls. A 2mg/mi collagen gel was
injected into the pFD, allowed to polymerize, then subjected to cyclic stretch of different strains.
At 12% strain, collagen gel in the device remained well adhered to the PDMS walls of the pFDs
(Figure 2-4a). However, at larger strain (22%), the gel was clearly observed to detach from the
walls. All subsequent tests with EBs were performed with a maximum of 10% cyclic strain.
A second test was designed to ensure that the cells were capable of responding to the mechanical
stimulus. For this purpose, human microvascular endothelial cells (hMVECs) were cultured on
the microfluidic channel surface and 10% strain, 1Hz uniaxial cyclic stretch was imposed (Figure
2-4b). Under these conditions, endothelial cells have been well documented to align
perpendicular to the direction of strain (103). Observed alignment was similar to what has been
reported in literature, confirming that the pFD platform is capable of translating mechanical
forces into cellular responses.
(a)
12% Strain, 1Hz
22%4 Strain, 1Hz
'ydC Streti
"
(b) 1Hz,
10% strain, 24 hours
Figure 2-4: Validations of the Proper Functions of the Microfluidic Device and the Cyclic Stretch
Platform
Three validations of proper functions of the motorized microfluidic system. (a) assessment of
hydrogel detachment with cyclic stretch and gel remained adhered to the wall below 22% strain; (b)
hMVECs aligned perpendicular to 10%, 1Hz strain, as reported in the literature.
2.4.2.
Differentiated Embryoid Bodies Exhibit Distinctly Different Morphologies in
Microfluidic Environment vs. Conventional Well Plates
EBs were generated with a standard hanging drop assay with 2 days of cell aggregation in a
droplet format and 3 days in suspension. Afterwards, EBs were transferred to gelatin-coated well
plates or microfluidic devices. EBs adhered to the gelatin coated surfaces in both culturing
environments. In well plates, EBs remained circular and spread out uniformly (Figure 2-5a), and
in the microfluidic system, some EBs self-organized into complex 3D structures after 6 days
(Figure 2-5b).
(b)
(a)
Figure 2-5: Morphological Differences of Embryoid Bodies
(a) When embryoid bodies were seeded in well plates, they remained circular and cells spread out
uniformly from all directions. (b) When EBs were seeded in microfluidic devices, they rearranged to
form complex aggregate patterns. Scale bar: 500pm
2.4.3.
Microfluidic Environment Enhances the Differentiation of Embryonic Stem Cells
into Cardiomyocytes
To examine the effects of confinement in the pFDs on cardiogenesis, compared with
conventional culture plates, EBs were either seeded directly on the substrate - 2D - or embedded
in a three-dimensional
hydrogel -
3D (Figure 2-6a).
GFP-positive and contracting
cardiomyocytes were detected 48-72 hours after EBs were allowed to adhere to a substrate
(Figure 2-6b). Compared with well plates, the rate of increase in GFP was higher in both 2D and
3D seeding conditions (Figure 2-7). When EBs were cultured on 2D, the increase in GFP
expression persisted over time in pFDs while those on well plates reached a plateau after three
days. A higher rate of differentiation was also observed in pFDs as opposed to culture plates
when EBs were suspended in 2mg/ml collagen I hydrogel (Figure 2-7b).
Conventional
(b)
Microfluidic Devices
GFP1
Figure 2-6: Method of Differentiation and Representative Images of Differentiated Cardiomyocytes
Procedure of differentiation and representative images of differentiated cardiomyocytes: (a)
procedures of ESC differentiations, EB formations and seeding conditions (b) representative images
of GFP-positive areas of EBs over time. Scale bar: 200pm
EBs Seeded on 2D
(a)
S16
14
12C
C
.
*12
1-- -FD
-- C3-
Well Plates
UU
-@-iFD
---
Xo.
Well Plates
* D 108
S10*-N
EBs Seeded in 3D
C (b)
.0
14-
8.N
2
2
00
_
_
_
_
_
_
4
3
2
1
Number of Days of GFP expression
_
_
_
_
__0
4
3
2
1
Number of Days of GFP expression
Figure 2-7: Microfluidic Environment Enhances Differentiation
Enhanced differentiation was observed for EBs cultured in 2D and in 3D hydrogel. Asterisks
represent statistical difference (p<0.05).
2.4.4.
Effect of Bone Morphogenetic Protein 2
The enhanced differentiation observed in microfluidic devices compared with conventional wellplates led us to hypothesize that cell-secreted cardiogenic factors accumulate in pFDs. We chose
bone morphogenetic protein 2 (BMP2) to test our hypothesis since blocking BMP2 signaling has
been shown to inhibit cardiac differentiation (104,105). With the addition of BMP 2 neutralizing
antibody (20pg/ml), differentiation in the pFDs was markedly diminished to a level similar to
that in well-plates suggesting that BMP2 plays an important role in the intensification of
differentiation in ptFDs and restricted transport in pFDs may contribute to the enhanced
differentiation (Figure 2-8).
Effect of BMP2 on Differentiation
C
0
161
a) U~14
a
Cn
0-Control
-C-BMP2 Ab
-
*
aCD 12
~_ 10
N
5~
_
U-
8
-
42
0
1
2
3
4
Number of Days of GFP expression
Figure 2-8: Effect of BMP2 on Differentiation
Addition of BMP-2 neutralizing antibody (20pg/ml) reduced differentiation in pFDs. Asterisks
represent statistical significance (p<0.05) between the two conditions
2.4.5.
Uniaxial Cyclic Stretch Inhibits the Differentiation of Embryonic Stem Cells into
Cardiomyocytes
When exposed to 24hrs of uniaxial cyclic stretch, embryonic stem cells differentiated
significantly less as compared to the unstretched controls (Figure 2-9). The same phenomenon
was observed for EBs cultured in 2D and in 3D. Moreover, EBs that did not express GFP prior to
mechanical stretch failed to express GFP over time (data not shown). Note that the only
difference between the static and stretched ptFDs is the 24-hour uniaxial cyclic stretch on day 3
of adherent culture. pFDs were subsequently cultured statically in both cases. This suggests that
even short term mechanical stretch interrupts the differentiation process and results in long term
reduction of cardiogenesis.
EBs Seeded on 2D
(a)
c
*
a)
LL0
aU)
N
(D =
161412108-
-
@
- Control
*
-0-- Stretched
EBs Seeded in 3D
(b)
14
C
.2
0
(I,
K CL 12
X
- .
-U--
- Control
Stretched
C 10
/7.
U-
O~8
6
$6=
WL
4.
0
IL
v 20.
00
4
3
2
1
Number of Days of GFP expression
4
3
2
1
Number of Days of GFP expression
Figure 2-9: Uniaxial Cyclic Stretch Inhibits Differentiation
Uniaxial cyclic stretch inhibited ESC differentiation: in both 2D and 3D, ESCs had inhibited
differentiation. Asterisks represent statistical significance (p<0.05) between the two conditions.
2.5.
Discussions and Conclusions
Embryonic stem cells have been considered as a cell therapeutic means to replenish myocardial
infarction zones with functional cardiomyocytes (18,48,106). It is, however, difficult to delineate
the effect of the mechanical contraction of the heart on stem cell differentiation in vivo and many
challenges remain in the differentiation, integration and incorporation of stem cells into the host
tissue (49). In this study, we have developed a compliant pFD to investigate the effects of culture
dimension and mechanical stretch on the differentiation of murine embryonic stem cells into
cardiomyocytes in vitro. This ptFD is modified from existing designs, developed in our
laboratory, including three-dimensional hydrogel regions to provide a more realistic extracellular
environment for cells in vitro and fluid channels to provide proper nutrient and waste transport,
and gas exchange. It allows for the study of mechanical stretch, which has been implicated to be
important in many biological processes, including the pulsation of the blood vessels from heart
contractions to proper embryo development (95-97). With an EB assay, this study aims to
characterize the differentiation process in response to uniaxial cyclic stretch, with frequency and
amplitude similar to the human heart.
We found that mechanical stretch at 1Hz and 10% strain inhibited differentiation. This finding
confirms two similar studies where the cardiomyocyte differentiation from human ESCs were
interrupted with cyclic mechanical strain (81,107). This has implications on how to best utilize
embryonic stem cells for cell therapy.
However, one other in vitro study of cardiogenesis with mechanical strain presented opposite
findings (79,98,107). Saha et al. suggest that TGFP/activin/nodal pathway is responsible for the
inhibition of differentiation, and Schmelter et. al. demonstrated that the reactive oxygen species
pathway is activated to enhance cardiomyocyte differentiation when stretched (79,98). While
these two studies use the same system (Flexcell® FX-4000T) to impose equibiaxial stretch, they
are different in many ways, which may explain the differences in results - stem cell line species
(human vs. mouse), differentiation protocols (colonies vs. EB formation), coating ECM proteins
(MatrigelTm vs. collagen I) and cyclic strain frequencies (10 cycles/min vs. static). These two
studies along with ours illuminate the complexity of the differentiation process. The effect of
mechanical stretch on cardiomyocyte differentiation can be highly dependent upon the
experimental setup. In our study, we chose to start the investigation with strain rates similar to
what human cardiomyocytes experience in vivo intending to mimic the physiological
environment.
Our findings suggest that mechanical stimulation disrupts the differentiation process prior to the
expression of a-MHC, a late-stage marker for cardiogenesis and tagged with GFP in this study
(108). This is supported by two observations. First, GFP expression was never observed for GFPnegative EBs after stretch. Second, for EBs expressing GFP prior to stretch, the overall GFP
expression did not increase after stretch, as it did in control. This can be due to the disruption of
differentiation for cells yet to express a-MHC by the time of stretch stimulation.
The
combination of mechanical stimulation and fluorescent reporting system utilized here can further
be used to elucidate the effect of stretch on the time course of cardiogenesis.
Another unique aspect of our study is the investigation of stem cell differentiation into
cardiomyocytes in 3D. Most in vitro studies used commercially available systems involving a
flexible membrane on the bottom of a well plate. Three-dimensional environments however
resemble the in vivo conditions more closely. In our custom-built system, cells can be seeded in
different conditions -on the pFD channels (2D) or suspended inside a collagen gel (3D). We find
50
the general trend of cardiogenesis is similar in 2D and in 3D. Nevertheless, this versatile pFD
platform can be further used to study the influence of mechanical stretch on other biological
processes where 3D environments are more desirable.
In addition to the advantage of the versatility of cell seeding arrangement, pFDs inherently
enhance the differentiation of embryonic stem cells into cardiomyocytes. By estimating the
amount of medium per EB (50pl for pFDs and 300ptl for well plates,) we hypothesize that the
enhancement of differentiation in pFDs is due to the accumulation of cell-secreted biochemical
factors responsible for differentiation. To test this hypothesis, we chose to inhibit BMP-2 with a
neutralizing antibody and showed a marked reduction in differentiation, to a level similar to EBs
cultured in well-plates, suggesting that the enhancement may be a result of accumulation of
BMP2 in pFDs. The usage of a small culture dimension can be considered as an alternative to
the addition of exogenous chemicals or adjustments of EB size to enhance cardiomyocyte
differentiation (69,106,108,109). Inhibition of other cardiogenic factors can also be tested in this
system.
This study deploys a well-controlled three-dimensional microfluidic system to study the
cardiogenesis process of murine ESCs with an EB assay. The effect of culture dimension and
uniaxial strain were characterized. First, we demonstrated that a higher EB to media ratio in
pFDs led to enhanced differentiation, partially due to the accumulation of cell-secreted BMP-2.
Furthermore, uniaxial cyclic stretch at 10% strain and 1Hz was found to be inhibitory for
differentiation. These findings provided additional insights on biophysical factors with which
cardiogenesis could be enhanced or inhibited.
Chapter 3. Effect of Extracellular Matrix Protein and Surface Patterning on
Driving Maturing Phenotypes of Isolated Embryonic Stem Cell Derived
Cardiomyocytes during Extended Culture**
**Most
content of this chapter has been accepted for publication in Cellular and Molecular
Bioengineering
3.1.
Introduction
In the last chapter, the differentiation of embryonic stem cells into cardiomyocytes in embryoid
body based assays was characterized. In this chapter, the cardiogenesis process was further
characterized by examining the phenotypic changes of isolated cardiomyocytes derived from
embryonic stem cells under various biophysical stimuli. Three phenotypic indicators are assessed:
sarcomere organization, cell elongation, and percentage of binucleation. As in the last chapter,
murine embryonic stem cells were differentiated in a hanging drop assay and cardiomyocytes
expressing GFP-a-actinin were isolated by fluorescent sorting. First, the effect of collagen I was
investigated. Addition of soluble collagen I markedly reduced binucleation as a result of an
increase in cytokinesis. Laden with a collagen gel layer, myocyte mobility and cell shape change
were impeded. Second, the effect of cell alignment by microcontact printing and nanopattern
topography was investigated. Both patterning techniques induced cell alignment and elongation.
Microcontact printing of 20pm line pattern accelerated binucleation and nanotopography with
700nm ridges and 3.5pm grooves negatively regulated binucleation. This study highlights the
importance of biophysical cues in the morphological changes of differentiated cardiomyocytes
and may have important implications on how these cells incorporate into the native myocardium.
3.2.
Background
The ability of embryonic stem cells to differentiate into cardiomyocytes in vitro has motivated
numerous studies of possible clinical application (48,110,111), development of in vitro model
systems (56,112,113), and/or generation of a continuous source of contractile cells. Some
challenges remain, however, such as how to drive embryonic stem cell derived cardiomyocytes
(ESCDMs) into maturation (114-116). In this chapter, we attempt to characterize the phenotypic
changes of ESCDMs with respect to two biophysical stimuli - presence of collagen I and cell
alignment. In Section 3.2.1 and 3.2.2, two environmental cues from the heart are closely
examined to see how they affect maturation. Then phenotypic maturation indicators, used in this
study, are described in Section 3.2.3.
3.2.1.
Collagen I in the Myocardium
Collagen I is a critical ECM component in the adult myocardium and a ubiquitous cardiac tissue
engineering scaffold material (3,55,117). The integrin interactions of collagen I in cardiac tissue
have also been well studied (10,118,119). For highly contractile cells, an overlay of collagen I
scaffold has been demonstrated to enhance differentiation and increase contractile protein
production in vitro (120). Some preliminary evidence suggests that by injecting collagen I
hydrogel with neonatal myocytes into infarcted tissue, cardiac ejection fraction is improved (121).
However, more mechanistic studies are required to understand how ECM proteins, collagen I in
particular, impact ESCDM function (21,112,121). However, more mechanistic studies are
required to understand how ECM proteins, collagen I in particular, impact ESCDM
function.(21,112,121).
3.2.2.
Effect of Surface Patterning on Cardiomyocytes and Cardiogenesis
The second biophysical stimulation we focused on is cell alignment. In the myocardium,
cardiomyocytes are arranged into layered two-dimensional sheets, precisely oriented to produce
a twisting motion during each contraction (122,123). Two ways to induce alignment have been
explored in vitro - protein patterning, also known as microcontact printing, and topographical
patterning. Both patterning techniques can induce anisotropy of neonatal myocyte monolayers
(53,60,61,124). However, the mechanisms by which patterns affect ESCDM phenotype remains
elusive.
3.2.3.
Maturation of Embryonic Stem Cell Derived Cardiomyocytes
Multiple types of cardiomyocytes have been studied in vitro: fetal cardiomyocytes, neonatal
cardiomyocytes, stem cell differentiated cardiomyocytes and adult cardiomyocytes, and each
type displays a different molecular, functional and phenotypic signature.
Phenotypically, binucleation increases as cardiomyocytes mature (125). More than 80% of the
cardiomyocytes in adult rats have more than one nucleus. Although the exact mechanism
remains elusive, it is speculated that as cardiomyocytes mature, they lose the capability to
complete cytokinesis and thus are unable to proliferate. The amount of binucleation has been
used as an experimental readout for maturing cardiomyocytes in vitro (126).
Another phenotypic indicator for maturation is the development of organized sarcomeres. Snir et
al. observed sarcomere development in human embryonic stem cell derived cardiomyocytes in
an embryoid body assay and found that randomly-distributed nascent myofibrils could be seen 7
days post seeding and well organized myofilaments were only found after 27 days (114). This
suggests that the organization of sarcomeres can also be used as a maturation indicator.
............
Furthermore, not only do sarcomeres become more organized over time, but they also increase in
width (127,128). During myofilament development, more myofibrils are recruited increasing the
length of z-disks in a contractility dependent fashion.
The contractile forces of cardiomyocytes also vary depending on the stage of development and
substrate stiffness (129). The optimal substrate stiffness to achieve maximum forces is found to
be similar to the stiffness of the native myocardium. From single cell force measurements, it has
been found that adult cardiomyocytes generate more forces than neonatal or fetal cardiomyocytes
(5.7pN vs. 3pN) (130,131). The difference in contractile forces has important implications for
the development on miniature actuators with cardiomyocytes.
Another phenotypic maturation indicator for developing cardiomyocytes is the changes in cell
shapes. As illustrated in Figure 3-1, cardiomyocytes elongate over time and adapt a more rodshaped morphology. The aspect ratio (ratio of major and minor axes) of cardiomyocytes
increases from 1 for fetal cardiomyocytes to 7 as adult cardiomyocytes. Cells also increase in
size (132). Jonker et. al. illustrated a significant increase in cell size during maturation for sheep
fetal cardiomyocytes (132). This increase in cell size and length reflects the finding of the
increase in number of sarcomeres in mature cardiomyocytes, as discussed previously.
f12d
n 6d
n12d
adult
Figure 3-1: Cardiomyocytes at Various Developmental Stages
Representative images of the change in cell shape at various developmental stage: fl2d, fetal
cardiomyocytes, n6d, early stage neonatal myocyte, nl2d, late neonatal cardiomyocyte and adult
55
cardiomyocytes (133). Cells elongate over time and become more rod-shaped they mature. Scale bar:
10pm.
In addition to changes in phenotype and force generating capacity, cardiomyocytes at various
developmental stages also exhibit molecular and functional differences. Functions of the
sarcoplasmic reticulum, calcium handling, and the localization of ryanodine receptors remain
immature for fetal cardiomyocytes and stem cell derived cardiomyocytes resulting in delayed or
diminished Ca
sparks (116,133-135). Other biochemical shifts, including changes in myosin
isoforms, are also important for the development of cardiomyocytes (136). These biochemical
changes can be detected with RT-PCR, western blot and/or immunofluoresence. The
biochemical variations and progressions are outside of the scope of this study and we focus
primarily on the phenotypic changes associated with maturation process.
3.3.
Materials and Methods
3.3.1.
Embryoid Body Dissociation and Fluorescent Activated Cell Sorting
Maintenance of undifferentiated murine embryonic stem cells and the formation of embryoid
bodies can be found in the Materials and Methods section of the previous chapter (Sections 2.3.1
and 2.3.2) Cardiomyocytes from EBs were isolated with fluorescent-activated cell sorting after
EBs were cultured on an adherent substrate for 96 hours (Figure 3-2). Although we found that
more cells differentiated over time, we chose to sort cells after 96 hours of adherent culture
because of the difficulty to dissociate the cells after a long period of time and rapid proliferation
of other cell types.
EBs are composed of stem cells with extensive extracellular matrix proteins and thus standard
trypsinization alone was not sufficient. Several common dissociation strategies were tested. Most
EBs remained as aggregates after incubation with collagenase for 30 minutes, the amount of time
we felt comfortable to leave EBs without medium. Vigorous pippetting resulted in cell lysis.
Lastly, passing cells through an 18G needle resulted in significant amount of cell death. We
finally optimized the dissociation technique with a two-stage trypsinization procedure.
EBs were first washed with PBS twice and incubated with 0.05% trypsin for 5-7 minutes. After
that, EBs were dislodged from the culturing substrate with gentle agitation using a 1ml pipettor
and transferred into a test tube where ample media were added to neutralize the effect of trypsin.
At this point, many clumps were still present and the test tube was centrifuged for 5 minutes at
1,200rpm. After the removal of the supernatant, more trypsin was added and the cells incubated
for another 5-7 minutes. This second stage caused most of the cells to dissociate. Medium was
added and the solution was centrifuged again. Once all trypsin-containing media were replaced
by fresh media, the content was passed through a 40pm cell strainer and transferred to a test tube
for sorting. Fluorescent activated cell sorting was performed with MoFlo (Cytomation). Either
differentiated ESCs without GFP or undifferentiated ESCs were used as the negative control. 2-4%
of total cells were GFP-positive and the purity of the sorted population was consistently higher
than 99.95%.
(b)
(a)
104
R3
s
10
Isolated 8 days after
hanging drop without
ascorbic acid
101
100
~2%
0V0801@aid
on
t
10~~~~~~
GFP Log
Figure 3-2: Fluorescent Sorting Results
(a) A typical fluorescent sorting graph. Undifferentiated ESCs were used as the negative
control/gating. Cells on the right hand side of the gate (GFP-positive) were collected. The purity was
consistently higher than 99.95%. (b) Optimization of isolation procedure. When we waited 14 days
before isolating ESCDMs, they were overtaken by proliferation of other cell types and difficult to
isolate from the extensive matrix. Addition of ascorbic acid improved the percentage of
cardiomyocytes harvested. Therefore, we cultured EBs with ascorbic acid and underwent cell sorting
8 days after the initiation of differentiation.
As an alternative to high-purity fluorescently-based sorting, Percoll gradient separation was also
tested to enrich the cardiomyocyte population from the ESC differentiated cell population.
Percoll gradient separation relies on the different densities of various cell types, cardiomyocytes
being heavier than other cells. We adapted a method described by Guo et. al. (56) and the
specific protocol can be found in Appendix II. According to Guo et. al. and our own findings, the
percentage of cardiomyocytes were enriched from 2-4% to 10%. This level of enrichment was
not enough to reach the high purity to study the phenotypic changes of ESCDMs. Thus, we
decided against using Percoll gradient as a viable option.
3.3.2.
Cell Seeding
The purified population of cells was resuspended in warm medium to a concentration of
2,000,000 cells/ml. Four collagen conditions were examined: ESCDM monolayer with no
collagen (control), ESCDM monolayer supplemented with 50pg/ml of soluble collagen I,
ESCDM monolayer laden with 2mg/ml collagen I hydrogel and isolated ESCDMs suspended in
2mg/ml collagen I hydrogel. ESCDM monolayer seeding density was 100,000 cells/cm 2, due to
the limited ability for differentiated cardiomyocytes to divide and proliferate. In the soluble
collagen condition, the small amount of collagen did not affect the pH of the medium. In the
condition with a collagen gel overlay, 2mg/ml collagen gel, prepared on ice, was reconstituted by
combining lOx DMEM, 0.5N NaOH, high concentration (3-4 mg/ml) collagen I (BD
Biosciences) and sterile cell culture grade water. 30pl of gel solution was used for each 96 well
plate, resulting in a hydrogel layer several hundred microns thick. The gel was polymerized for
40 minutes in the incubator (37'C, 5% CO2), followed by the addition of warm medium. The
scaffold was added to the cell monolayer three days after seeding, when most cells had fully
spread and formed confluent monolayers. For the condition with ESCDMs suspended in collagen
gel, the same collagen recipe was followed except that water was replaced with cell solution for
a final concentration of 1.5x106 or 5x106 cells/ml. Schematics of the seeding conditions can be
seen in Figure 3-3.
Control - no addition of ECM
SolCol - 50ptiml collagen I in media at day 3
(AlGel - 2mg'ml colla en I gel laden on monolayer at day 3
Singlec(lls mix single cells in 2mg/nil collagen I gel
Figure 3-3: Schematics of Cell Seeding Conditions
Four different ESCDM seeding conditions are tested to study the effect of collagen I on the
maturation of ESCDMs - control, soluble collagen, collagen gel laden on top of an ESCDM
monolayer and single cell suspensions.
3.3.3.
Microcontact Printing (MCP)
Microcontact printing is a technique to transfer specific proteins between two solid surfaces. The
procedure used is similar to that described in Feinberg et. al. (53).
Briefly, molds were
fabricated by standard soft lithography and stamps with a 20pm/20pm line pattern were made
with polydimethylsiloxane (PDMS). Stamps were sterilized by sonication in 50% ethanol for 30
minutes and stored in 70% ethanol. Stamps were washed, dried and incubated with 0.1%
Oregon-green conjugated gelatin (Invitrogen) for 1 hour. Rinsed and dried stamps were lowered
onto plasma treated surface to initiate contact and pressed for 1 minute with mild pressure
allowing for proper protein transfer.
In the following steps, one rinse with sterile PBS was performed between each step. 1%
Pluronics-F127 (Sigma) was added to the stamped and rinsed surface for 15 minutes, followed
60
by incubation of 0.02% nonfluorescent gelatin (Gibco) for 15 minutes.
Finally, the surfaces
were rinsed three times before being stored in PBS in the incubator.
3.3.4.
Nanopattern Topography (NPT)
Hot embossing generated the submicron-scale substrate topography with an epoxy mold created
by a three-step molding process. First, a silicon master mold was fabricated using standard
photolithography and reactive ion etching (RIE). A 1 pm-thick layer of silicon oxide was
thermally grown on a 100 mm silicon wafer and features ranging from 500 nm to 1 pm were
created by spin-coating a 500 nm layer of Shipley 1805 photoresist on the wafer, patterning the
photoresist using a standard photomask, and developing the patterned photoresist. An anisotropic
reactive ion etch using CF 4 and 02 transferred the photoresist features into the silicon oxide to a
depth of 1 pm before stripping the photoresist. The second mold was made with PDMS. The
third epoxy mold was made from a two-part high temperature epoxy (Conapoxy FR-1080,
CYTEC Industries). For the hot embossing step, the epoxy mold was placed with the submicron
features in contact with a thermoplastic polystyrene blank in a uniformly heated, temperatureand pressure-controlled press. The resulting substrate was a relief replica of the silicon master
mold and served as the platform for cell culture experiments.
3.3.5.
Immunofluorescent Staining
To visualize sarcomere organization, cells were first washed twice with 1x PBS to remove media
and fixed with 4% paraformaldehyde (PFA) for 15 minutes in room temperature. 0.1% triton-x
was used to permeabilize the cell membrane for 2-5 minutes. For the following steps, 2D
cultures were incubated with the appropriate reagent for 1hr at room temperature, and 3D
cultures were incubated overnight at 4'C. Cells were first incubated with block ace (Dainippon
Pharmaceutical, Tokyo, Japan) to block nonspecific binding, followed by incubation of primary
61
I'll,
................
...
antibodies, mouse sarcomeric a-actinin, at dilution factor of 1:200. Secondary detection antibody
(anti-mouse Alexa 568) and 4',6-diamidino-2-phenylindole (DAPI) were used to visualize
protein of interest and nuclei respectively. In between each of the steps described previously, two
rinses of 1x PBS were performed. For samples with collagen gel, longer washing procedures
were adapted and secondary antibody was reconstituted in PBS with 2% rabbit serum and 2%
bovine serum albumin to further reduce background staining.
3.3.6.
Quantification and Statistical Analysis of Binucleation, Cell Shape and Cell
Alignment
Cell elongation, binucleation and cell orientation were quantified by ImageJ. To quantify
elongation, individual GFP-positive cells were identified, traced and fitted to an ellipse. Then the
ratio of major and minor axis, the aspect ratio, was calculated. For each condition, 10-30 separate
regions (330ptm x 430pm) were imaged and a total of 60-100 randomly selected cells were
traced. Thus, the majority of the cells traced were not in contact with each other. The minute
difference in GFP-a-MHC expressions between cells, the cell outline could be visualized. This
was validated and compared against an ESCDM monolayer stained with N-cadherin (Figure 3-4).
Figure 3-4: Cell Boundary Identification and Verification with N-Cadherin
To quantify binucleation, each GFP image was merged with DAPI image to enhance the
identification of nuclei. Number of binucleated cells were then counted. At least 500 GFPpositive cells were counted for each experimental condition. To measure alignment angle, pattern
angle was first determined and then the deviation angle of the major axis of the cell and the
pattern angle was calculated. At least three independent sets of experiments were performed for
each condition.
Data were expressed as mean ± SEM and analyzed using one-way ANOVA followed by post
hoc Student's t test. Results with P<0.05 were considered statistically significant.
...........................
3.4.
Results
3.4.1.
Sarcomere development of ESCDMs suspended in collagen gel or plated on
ECM-coated polystyrene
In this study, we were interested in how biophysical stimuli affect the morphology of ESCDMs.
We started the investigation with collagen I, a ubiquitous ECM protein in the adult heart as well
as in previous in vitro cell culture with cardiomyocytes (55,56). We found that suspending
ESCDMs in 2mg/ml collagen gel does not induce sarcomere development or cell-cell contacts,
two key parameters observed during proper cardiomyocyte development (Figure 3-5) (137-139).
Isolated ESCDMs remain rounded and contractile. Immunofluorescent staining of sarcomeric aactinin reveals a punctate protein distribution, lacking any organized structures. Two cell seeding
densities (1.5x 106 and 5x10 6 cells/ml) exhibited similar results. These findings suggest that static
culture of single ESCDMs suspended in collagen I gel is not conducive for myocyte
development, at least for the times tested in these experiments.
Nucleus
( -M(
(c) Sarc, aj
acti n
'(e) 1
d
eg
i,51(f 5.rr/m
Figure 3-5: ESCDMs Embedded in Collagen I Hydrogel
....
........
...........
......
..
...........
.......
ESCDMs are suspended in 2mg/mi collagen I hydrogel immediately after fluorescent sorting. They
fail to induce cell-cell contacts or sarcomere formation. Sarcomere a-actinin distribution remains
punctuate (e), and cells remain viable, isolated and rounded (a)-(d), suggesting that suspending
ESCDMs as single cells in collagen gel does not promote cardiomyocyte development. These cells are
fixed after 21 days.
Scale bar for (a)-(d): 100pm; (e): 10pm.
In contrast, confluent ESCDM monolayers with well-developed sarcomeres, along with stress
fibers, are observed in cells seeded on ECM-coated polystyrene. Sarcomeres, albeit immature,
can be observed 48 hours after seeding and become more clearly evident for all three conditions
-
no collagen, with soluble collagen and with a laden layer of collagen gel (Figure 3-6).
Day 2
Day7
Day 14
Figure 3-6: Representative Images of ESCDMs on Polystyrene
Representative images of ESCDMs cultured on gelatin coated polystyrene on day 2, day 7 and day 14.
In contrast with ESCDMs suspended in collagen, sarcomeres, visualized with sarcomeric a-actinin, is
readily observed as early as 2 days after seeding. Confluent and mature sarcomeres can be most
clearly observed at Day 14. Scale bar = 50pm
3.4.2.
Cell Adhesion with Overladen Collagen Gel Layer
One experimental condition involved overlaying a layer of collagen gel of several hundred
micrometers on top of the ESCDM monolayer on day 3 of adherent culture. To examine whether
ESCDMs were bound to the overlaying collagen gel or to the tissue culture substrate, we gently
removed the gel from the wells and examined the gel and the wells separately (Figure 3-7).
ESCDMs were mostly adhered to the well plate 7 days after seeding (4 days after introducing
collagen gel,) but at two weeks after seeding, most ESCDMs were bound to the gel.
Similar study has been conducted by Vandenburgh et. al. where myotubes exhibited enhanced
differentiation and attached firmly to the collagen overlay on top (120).
Figure 3-7: Cell Adhesion to Overlaying Collagen Gel
Most ESCDMs stayed adhered to well plates on day 7, but transferred to attach to the overlaying
collagen gel after 14 days.
3.4.3.
Effect of Collagen I on the Elongation of ESCDMs
Elongation of ESCDMs was measured by the aspect ratio - the ratio of the major and minor axis
-- of each cell. Figure 3-8a shows the aspect ratio of ESCDMs over time, exposed to collagen I in
different forms. Regardless of how collagen I is presented to the cell monolayer, cell aspect ratio
increases over time. The aspect ratio of ESCDMs laden with a layer of collagen I is reduced
slightly but significantly compared to control where no additional collagen is added.
66
Interestingly, the rate of increase in aspect ratio is dependent upon the collagen condition (Figure
3-8b). Clear increase in aspect ratio over time can be observed for the three different collagen I
conditions but the increase of aspect ratio is highest for the soluble collagen condition, followed
by the control condition and the collagen gel condition. On average, the aspect ratio of ESCDMs
increases by 9.6% per day for the condition with soluble collagen, 7.5% for control, and 4.6% for
the condition with collagen gel.
(a) Aspect Ratio of ESCDMs with Various ECM Conditions (b) 2.4
Day2
Day2-SolCol
4
tDay7
2.23.5TW
Normaltzed Increase InAspect Ratio
control
aCoIGeI
Day14
3.5
1
*
3
2.5
1.5
1.4
05
Ctrl
Solcol
ColGel
4
6
8
1
2
1
Figure 3-8: Effect of Collagen I on Aspect Ratio
(a) Aspect ratio of ESCDMs increases over time regardless of the condition of collagen gel. Asterisks
represent statistical difference (p<0.05) of the data point of interest compared to day 2 of the same
condition and diamond (0) stands for statistical difference (p<0.05) of the data point of interest
compared to the control condition without additional collagen I of the same day. (b) There is a
difference in the rate of aspect ratio increase for different collagen conditions. ESCDMs cultured
with soluble collagen elongate most rapidly, followed by control and collagen gel.
3.4.4.
Effect of Collagen I on Binucleation
Binucleation, like cell elongation measured by aspect ratio, increases over time (Figure 3-9a).
Binucleation has been suggested to be a phenotypic indicator of maturation in vivo and in
vitro.(125,126) Effects of collagen I on binucleation depend on how collagen I is presented.
Presented as a hydrogel, it has little impact on binucleation. A small increase in binucleation can
be seen after 2 weeks of culture with collagen gel, compared with control. When collagen I
solution is added into the medium, it strongly inhibits binucleation. After 14 days, ESCDMs
cultured with soluble collagen have 50% fewer binucleated cells compared with the control.
To confirm whether the reduction of binucleation is due to a decrease in karyokinesis or an
increase of cytokinesis, we counted the number of cells. We found an increasing number of cells
when ESCDMs are cultured with soluble collagen I, under which condition ESCDMs are more
likely to undergo cytokinesis and complete the cell cycle (Figure 3-9 b).
(b)
(a)
20
01
C')
01
*
ECtri
ESolCol
500
E
E 400
5 EZColGel
0*
0
C,)
S300
0
C,
200
5
E
: 100
z
0 Day2
Day7 Dayl 4
OL
Day2
DDa11
4
Day7 Dayl 4
Figure 3-9: Effect of Collagen I on Binucleation
Effect of collagen I on binucleation is presented. (a) ESCDMs are either exposed to no collagen,
soluble collagen (50pg/ml) or 2mg/mI collagen gel laden on top of the monolayer seeded on an
isotropic surface. (b) Number of cells with different collagen. Asterisks represent statistical
difference (p<0.05) of the data point of interest compared to day 2 of the same condition and
diamond (0) stands for statistical difference (p<0.05) of the data point of interest compared to the
control condition without additional collagen I of the same day.
The effect of collagen I on patterned surface can be observed in Figure 3-10. Similar to the
trends on isotropic surface, the percentages of binucleated cells increase over time. Microcontact
printing does not affect the general behavior of binucleation. However, nanopattern topography
surface is a negative regulator (more details in section 3.4.7). ESCDMs cultured on NPT surface
with soluble collagen have markedly reduced amount of binucleated cells.
(a)
(b)
20.0%
*
18.0%
16.0%
14.0%
12.0%
1.0%
8.0%
6.0%
4.0%
?.0%
2.0%
20.00%
18.00%ontrol
n/SolCol
ColGel
1600%
*
14.00%
12.00%
10.00%
800%
6.00%/
4.00%
2.00%
0
-
0.00%
Day
Day 7
Day 14
Day 2
Day 7
Day 14
Figure 3-10: Effect of Collagen I on Binucleation on Patterned Surfaces
Percentages of binucleated cells on (a) MCP and (b) NPT surfaces are recorded. Asterisks represent
statistical difference (p<0.05) of the data point of interest compared to day 2 of the same condition
and diamond (0) stands for statistical difference (p<0.05) of the data point of interest compared to the
control condition without additional collagen I of the same day.
3.4.5.
Effect of Surface Patterning on ESCDM Alignment
The second biophysical cue we investigated is cell alignment. ESCDMs cultured on MCP and
NPT surfaces align with the pattern surface (Figure 3-11). For the MCP surface, gelatin
conjugated with Oregon Green is used to visualize and measure the pattern angle. For the NPT
surface, the ridges and grooves alter between 3.5pm and 700nm. Aligned sarcomeres can be
69
........................
observed for both patterning techniques. Sarcomere development is strongly dependent on the
periodicity of NPT surface as the sarcomere width coincides with the width of the ridges. This
same phenomenon is not observed for MCP surface.
Microcontact
Printing
(M rP\
Nanopattern
Topography
(NPTI
(a)
Patterns
(b)
20x
View
(C)
80x
View
1 OPM
Red: Sarc. a-actinin Blue: Nucleus
Figure 3-11: Cell Alignment with Microcontact Printing and Nanopattern Topography
(a) Two methods of patterning were tested - microcontact printing (MCP) which distributes proteins
of alternating concentrations onto a substrate in a line pattern, and nanopattern topography (NPT)
which has nano-scale grooves and ridges where cells bind on top. MCP was visualized by stamping
gelatin conjugated with Oregon green. Scale bar = 100pm.
(b&c) Cells exhibit clear alignment when cultured on MCP and NPT surfaces. These are
representative images of nuclei (DAPI), sarcomeric a-actinin and merged cells from day 14 samples
with 50pg/ml soluble collagen. Scale bars: (b) 50pm, (c) 10pm.
.......
.....
For cells cultured on MCP and NPT surfaces, the angle between the major axis of the cell and
the pattern was calculated (Figure 3-12). Alignment can be observed within 48 hours: almost all
cells cultured on MCP and NPT surfaces align within 20 degrees of the pattern orientation
(Figure 3-12a). The same strong alignments persist over time (Figure 3-12b&c). By day 7, more
than 70% of the cells cultured on NPT surfaces align within 10 degrees from the pattern.
Interestingly, cell alignment on MCP surface is better retained with collagen gel overlay after 2
weeks. 60% of cells in that condition remained aligned within 10 degrees while alignment in
other conditions declined to 30% (Figure 3-12d).
00
(a)
Isotropic
MCP
75
(b)
0
MNPT
0
a) 50
CID
25
as
2~ -90-70-50-30-10
100
0)
CM
10 30 50 70 ,
1
angle (deg)
M MCP-SolCol
M NPT-SolCol
50
Isotropic-Gel
MCP-Gel
-
Isotropic-SolCol
M MCP-SolCol
50
-
NPT-SolCol
Isotropic-Gel
MMCP-Gel
25-
NPT-Gel
-10-70-50-30-10 10 30 50 70 90
angle (deg)
0-0
Isotropic-Otrl
MCP-Ctrl
NPT-Ctrl
Isotropic-SolCol
100
00NPT-Ctri
AL||1. . , i
a)
(d)
Isotropic-Ctrl
MCP-Ctr
75
25
C
,
00
0
a)
CM
Ctr
SolCol
2
50-
MColGel
NPT-GeI
a) (
-k0-70-50-30-10
10 30 50 70 90
a. cCD0
0
MCP
angle (deg)
angle (deg)
NPT
Figure 3-12: Histogram of Cell Alignments
Histograms of cell alignment at (a) day 2, (b) day 7, and (c) day 14 are plotted. Three surface
patterns are compared: no patterns, MCP, and NPT. Three extracellular matrix incorporations are
compared for day 7 and day 14 (the conditions do not apply to day 2 samples because they are
applied on day 3): no additional ECM, 50pg/ml soluble collagen I in media and 2mg/ml collagen
hydrogel laden on the surface of the monolayer. Because of the secretion of ECM by the cells, we
tested if the alignment would be maintained on MCP surface at day 14. (d) illustrates that while
ESDMs cultured without collagen and with soluble collagen lose alignment, alignment for ESCDM
monolayers laden with collagen gel is rescued.
Effect of Surface Patterning on ESCDM Elongation
3.4.6.
Cells cultured on patterned surfaces, both MCP and NPT, exhibit elongated features (Figure 3-11)
although the rate of elongation is highest for cells cultured on isotropic surfaces (Figure
3-13a&b). There is virtually no increase in aspect ratio for cells on MCP surfaces suggesting that
the cellular response to the surface has saturated after 48 hours.
Elongation is mostly attributed to a decrease in minor axis length rather than an increase in major
axis length (Figure 3-13 c&d). Significant reductions in minor axis are observed for both the
isotropic and NPT conditions and this explains the shrinkage of the monolayer away from the
walls at longer time points.
(a)
(b)
8
6
0
.-
KKK
4
2
-n-Isotropic
'
r 1 -,-MCP
O NPT
0
Cl)
Isotropic MCP
10
5
Time (Days)
0
NPT
15
(d)
(c)
0 00
0 Isotropic
MCP
-
Ii
NPT
E
40
**
i 20
U)
-J
Isotropic MCP
NPT
Figure 3-13: Effect of Surface Patterning on Elongation without the Addition of Collagen
Effect of surface patterning on cell elongation is measured. (a,b) Aspect ratio of cells on isotropic and
NPT surfaces increases over time while that of cells on MCP surface remains constant. (c,d) The
average dimensions of the major and minor axes are also recorded and it can be seen that most
increase in aspect ratio is attributed to a decrease in minor axis. Asterisks represent statistical
difference (p<0.05) of the data point of interest compared to day 2 of the same condition and
diamond (0) stands for statistical difference (p<0.05) of the data point of interest compared to the
isotropic condition of the same day.
The aspect ratios of ESCDMs in the presence of soluble collagen or collagen gel are shown in
Figure 3-14. In the presence of soluble collagen, ESCDMs cultured on isotropic and NPT
surfaces elongate over time and ESCDMs cultured on MCP surface do not clearly elongate.
When ESCDM monolayer is laden with a layer of collagen gel, ESCDMs on isotropic and NPT
surfaces clearly elongate and a less significant elongation is observed for ESCDMs on MCP
surface.
*
(a)
(b)
8
8
6
6
4
4
2
2OI
Day14
0
Isotropic
MCP
VDay7
0
0
Isotropic
(a
MCP
NPT
NPT
Figure 3-14: Elongation of ESCDMs with Soluble Collagen and Collagen Gel
Effect of surface patterning on ESCDM elongation with (a) soluble collagen or (b) collagen gel is
documented. Asterisks represent statistical difference (p<0.05) of the data point of interest compared
to day 2 of the same condition and diamond (0) stands for statistical difference (p<0.05) of the data
point of interest compared to the isotropic condition of the same day.
3.4.7.
Effect of Surface Patterning on ESCDM Binucleation
Percentage of binucleation increases over time for all patterning conditions. Binucleation rate
between day 2 and day 7 of ESCDMs on MCP surface is much higher than in cells on isotropic
surfaces (Figure 3-15a). Markedly reduced binucleation was observed for ESCDMs on NPT
surfaces, compared with isotropic surface regardless of presence of collagen gel (Figure 3-15b).
We found an increase in the number of ESCDMs on NPT surfaces indicating that the reduction
in binucleated cells is a result of cells completing the cell cycle and being able to undergo
cytokinesis (Figure 3-15c). This is similar to the finding for ESCDMs cultured with soluble
collagen.
r(a)30
a)20
(b)13
Day2
Day7
*
*
Day14
CY)
Day7/Day2
(C)
Ctrl
140%
600
oj-
(D 20 M Iso-ColGel
CM
NPT-ColGel
a10
Q10
00
Iso-CtrI
NPT-CtrI
MCP
266%
Gel MCP-Gel
132%
338%
Day 2
Day 7
Day 14
isotropic
MCP
NPT
5 E 400
E
-
.*200
z
0
Day2
Day7
Dayl4
Figure 3-15: Effect of Surface Patterning on Binucleation
Binucleation is a sign of maturation as it is widely observed for adult myocytes but rare in neonatal
myocytes. (a) MCP accelerates binucleation between day 2 and day 7. (b) NPT is a negative regulator
of binucleation. (c) By examining the number of cells, there are a higher number of cells on NPT
surface after 14 days suggesting that the decrease in binucleation is a result of an increase in
cytokinesis. Asterisks represent statistical difference (p<0.05) of the data point of interest compared
to day 2 of the same condition and diamond (0) stands for statistical difference (p<0.05) of the data
point of interest compared to the isotropic condition of the same day.
3.5.
Discussion and Conclusions
In this study, we explored the effects of cell alignment and the addition of extracellular matrix
protein, collagen I specifically, on changes of embryonic stem cell derived cardiomyocytes
phenotype. This work is motivated by two observations. First, due to the limited regenerative
ability of the heart (140), differentiated stem cells have been considered as a possible source of
functional cardiomyocytes. They can also be used as in vitro model systems (141,142). Second,
spontaneously contractile cardiomyocytes can be used as miniature actuators. To date, most
research focused on biomachines utilizes neonatal myocytes (50,52,143). Stem cell derived
cardiomyocytes have the advantage of being an unlimited and continuous source of contractile
cells. However, little is known about how biophysical factors affect ESCDMs and we attempt
here to gain a better understanding by first probing two stimuli inspired by the natural
environment of the myocardium as well as the previous literature.
We have portrayed a rather complex picture of the relationship between collagen I and ESCDMs.
We chose collagen I because of its abundance in the adult heart even though other ECM proteins
are present at various stages during development (3). It is also ubiquitously used as a tissue
engineering scaffold in vitro (120,144). First, we observed that by suspending ESCDMs in
collagen as single cells statically, no sarcomeres or cell-cell contacts are formed, even at a high
cell density. This result is similar to the findings from Souren et. al. where neonatal
cardiomyocytes embedded in collagen gel remain isolated and contract asynchronously (145).
The lack of adequate growth factors or matrix proteins may be one explanation for this finding
since Zimmermann et. al. successfully developed 3D cardiomyocyte construct with the addition
of MatrigelTM and chick serum (55). Another possible explanation is that the stiffness of the
hydrogel, estimated around 10 KPa (146), is softer than optimal for myogenesis (147-149). On
75
the other hand, according to Discher et. al., substrates that are too stiff, such as polystyrene, are
suboptimal for myocyte development as well (150). However the stiffer substrate is better able to
withstand the traction force required for sarcomere development (151,152), and this may provide
another possible explanation for the differences observed.
Secondly, cell elongation is impeded and alignment maintained in the presence of collagen gel
overlay. (Figure 3-8a) This might be due to the physical constraint the hydrogel imposes on the
monolayer, perhaps by binding to the monolayer, and/or by the weight of the gel. One study has
found that a collagen gel overlay improved the differentiation of skeletal myotubes presumably
by imposing a constant load on the monolayer (120). Combined with MCP, the collagen gel
accelerates binucleation, as demonstrated by the significant increase in binucleation between
days 2 and 7, although the increase in binucleation between day 7 and 14 is not statistically
different from control (Figure 7a). Although the mechanism of this phenomenon remains to be
elucidated, this finding is one example of the combinatorial effects of two biophysical
stimulations on ESCDMs.
While this study focuses on the effects of collagen I, other ECM proteins, such as vitronectin or
fibronection, may be present in the experimental system due to the presence of serum in the
medium. These ECM-integrin interactions have been demonstrated to be critical during heart
development (65,117,153). In this study, variation due to the presence of serum was mitigated by
using the same lot of serum although further studies could be envisioned to specifically
investigate the effects of other ECM proteins on ESCDMs.
Lastly and unexpectedly, collagen I, presented in a soluble form in medium, is a strong negative
regulator of binucleation. This effect is not seen when it is presented as a hydrogel. Soluble
collagen increases cytokinesis and promotes ESCDMs to complete the cell cycle. This suggests
that how ESCDMs respond to collagen I depends on the way it is presented to the cells. This is
reminiscent to the differential responses of cells to exogenous and matrix-bound growth factors
and this finding is, to our understanding, the first study to explore the effects of exogenous
collagen I on ESCDMs (154).
The second biophysical stimulation we tested is the effect of cell alignment, induced by
microcontact printing and nanopattern topography, on ESCDM morphology. In this study, only
one set of geometric dimensions of each patterning techniques is explored. Bursac et. al. found
that varying MCP dimensions did not cause a statistically significant difference in the
electrophysiological properties of neonatal myocytes (59). For nanoscale topography, Au et. al.
found neonatal myocytes exhibited clearer sarcomere structures on surfaces with 0.5ptm-wide
ridges and 0.5pm-wide grooves than lpm-wide ridges and 3pm-wide grooves, although it is
unclear whether the difference arises from differences in duty cycle or the exact dimensions (60).
Due to the limited understanding of the effects of NPT and the difficulty to directly compare the
two patterning techniques (60,61), we focused on characterizing their effects on ESCDMs
separately compared with the isotropic control.
In this study, consistent with previous literature, we report that both patterning techniques induce
cell alignment and elongation, as observed in cardiomyocyte sheets in vivo (3). The increase in
aspect ratio over time is mostly attributed to a decrease in width and the lateral dimensions of the
cells remain significantly smaller than adult myocytes (although cell volumes were not measured)
(155). Longer culture times and mechanical/electrical stimulations may be necessary for
enhanced differentiation and maturation (114,156,157).
In terms of binucleation, the MCP surface alone, without the addition of collagen, slightly but
significantly increased the percentage of binucleation compared with a smooth surface. In
contrast, compared with the isotropic case, ESCDMs cultured on NPT surface observed
markedly reduced binucleation, due to increased cytokinesis. It is reasonable to believe that the
finding for MCP surfaces is less dependent on the geometric dimensions but it is less clear
whether varying NPT dimensions will produce the same finding. Nevertheless, from this study,
we are able to conclude that induced cell alignment affects the binucleation of ESCDMs in vitro.
Presence of collagen I and alignment are two important biophysical characteristics of the adult
myocardium and here we adopted these two stimuli and characterized their effects on ESCDM
phenotypic changes in vitro. We are able to demonstrate that by affecting them, we can alter the
morphology of embryonic stem cell derived cardiomyocytes. This is an important step in
advancing our understanding of how to utilize embryonic stem cell derived cardiomyocytes as a
cardiovascular disease treatment option. By being able to control the phenotype in vitro, this
opens the prospect of using ESCDMs as actuators for biomachines.
Chapter 4. Synchronization, Force Generation, and Control of Contractile
Direction of Embryonic Stem Cell Derived Cardiomyocytes
4.1.
Introduction
In the last two chapters, the cardiogenesis process is characterized. Specifically, the process was
studied in two stages - the differentiation of ESCs to cardiomyocytes and the subsequent
phenotypic progression of nascent cardiomyocytes. In this chapter, the focus is shifted from the
characterization of cell behaviors to applications of biomachines powered by ESCDMs.
ESCDMs, unlike neonatal or adult cardiomyocytes, have the potential to be an unlimited cell
source without handling and sacrificing animals. To utilize ESCDMs as an actuator, three
important attributes necessary for a coherent actuator were investigated - in-phase contraction
for the cell monolayer, force generation estimates and control over contractile directions.
To study in-phase contraction, cell contractions were estimated as the changes in cell lengths
over time with a MATlab-based algorithm. Cells at different areas of the monolayer contracted
and relaxed simultaneously suggesting that the entire monolayer was in mechanical unison.
Second, the amount of force the ESCDM monolayer generated was estimated by tracking
motions of fluorescent microbeads suspended in collagen gel which was laden on top of the
monolayer. With scaling analysis, ESCDM monolayer was found to generate forces that matched
well with the reported values for neonatal myocytes. Lastly, the direction of contraction was well
controlled with surface patterning techniques, described in the previous chapter. Techniques used
in prior studies to characterize ESCDM phenotypes were applied to advance our understanding
in how to utilize ESCDMs for non-medical applications and these three tests validated the
possibility to control and utilize aggregates of ESCDMs as actuators.
4.2.
Background
The contractile nature of cardiomyocytes lends itself to be an attractive candidate as an actuator.
Different types of machineries powered by cardiomyocytes have been developed. Neonatal
myocytes are the primary choice for these biomachines because, unlike adult myocytes, neonatal
myocytes do not lose their contractile phenotype in vitro. The biggest limitation in using
neonatal myocytes is the need to maintain an animal facility and sacrifice animals. ESCDMs
eliminate this limitation and have the potential as an unlimited source of myocytes.
There are two general approaches to cardiomyocyte-powered machines - seeding cells on 2D or
embedding them in a three-dimensional hydrogel mixture. Quantitative measurements and
details can be found in Table 1 in Section 1.3.
4.2.1.
Utilizations of Cardiomyocytes as Biomachines in 2D
Feinberg et. al. cultured neonatal myocytes on a thin PDMS film, printed with anisotropic
protein patterns (53). By cutting the thin films along the cell alignment angle or off by 5-15',
they were able to produce forward or twisting motions of the cell-seeded films, respectively.
Tanaka et. al and Kim et. al applied similar concepts to produce a spherical pump, a linear pump
and a three-legged walker by seeded with cardiomyocytes on 2D surfaces (50,52,158). All of
these biomachines require a structural support from a polymer such as PDMS.
Xi et. al. developed a system which required no PDMS support. The approach utilized a
thermally responsive polymer, Poly-(N-isopropylacrylamide), which provided the support during
seeding and could be shrunken away simply with heating (143). With this technique, they
developed a cantilever and a walker with cardiomyocytes. This study illustrates the possibility to
combine tissue engineering and material science to create a microscale device powered by
muscle.
The pumps developed by the Tanaka et. al. and Kim et. al. were examples of the integration of
sophisticated microfabrication techniques and cell culture. However, they have yet to utilize the
intrinsic advantages and behaviors of cardiomyocytes, such as cell-alignment with proper cues
and synchronization of cell monolayer. The adaptability and responses of cells to environmental
stimuli may result in superior bioactuator designs.
4.2.2.
Utilizations of Cardiomyocytes as Biomachines in 3D
A different approach to utilizing cardiomyocytes is to suspend them in a hydrogel and allow the
cells to assemble and form a tissue construct. This approach takes advantage of extensive
insights from tissue engineering. Engineered heart tissue (EHT) is comprised of a mixture of
cardiomyocytes, collagen I, MatrigelTM, chick serum, and medium (21,22). The cell-gel solution
was polymerized into a toroid shape and eventually assembled into a tissue construct. A similar
approach was taken by Guo et. al. (56). However, these constructs often did not exhibit coherent
contraction and the toroid shape does not resemble the native cardiac tissue. Furthermore, the
cell purity of these constructs is difficult to determine rendering systematic variations and
characterizations impossible.
4.3.
Materials and Methods
4.3.1.
Cell Culture and Isolation
Detailed protocols of undifferentiated stem cell maintenance and induction of differentiation can
be found in Sections 2.3.1 and 2.3.2 respectively. Isolation and seeding are described in Sections
3.3.1 and 3.3.2. All the experiments were performed with ESCDMs isolated with fluorescent
sorting.
4.3.2.
In-phase Contraction Analysis
ESCDMs were cultured on gelatin-coated tissue culture grade polystyrene at a density of
100,000/cm2 and a confluent monolayer was formed after 3 days. Contraction of the ESCDM
monolayer was captured with Zeiss Axiovert 200 microscope fitted with a Hitachi CCD color
camera with a capture rate of 30 frames per second.
Once a video was captured, it was imported into a Matlab-based particle tracking program based
on Polynomial Fitting with Gaussian Weight, capable of detecting particles with relatively low
contrast (159). Three cells on three separate regions of the image were chosen. For each cell, two
particles were selected on either side of the nucleus, as close to the cell edge as possible (Figure
4-1). The shortening (contraction) and lengthening (relaxation) of each cell were recorded as the
changes in distance between the two points.
Figure 4-1: Particle Tracking to Investigate Synchronization
To detect contraction across monolayer, three separate cells were selected and two particles on either
side of each cell were tracked. The change in length of each cell over time was recorded as the
distance between the two points. Scale bar: 100pm
4.3.3.
Force Generation
To estimate force generation, 2mg/mi collagen gel with 2pm red fluorescent beads at a
concentration of 2x106 beads/ml (Spherotech) was laden on confluent ESCDM monolayer
(Figure 4-2). The detailed description of this protocol can be found in Section 3.3.2. Cell
monolayer was detected via fluorescent microscopy with 488nm filter. Then the filter was
switched to 568nm to microbeads close to the monolayer. Movements of beads were recorded
with a fluorescent microscope and analyzed with the same particle tracking program described in
the last section. As a control, gels with suspended beads were cultured and polymerized without
the cell monolayer. In the control case, only Brownian motion could be observed for the beads
(not shown).
11
11
........
..:::
Figure 4-2: Fluorescent Beads to Estimate Force Generation
ESCDM monolayer (a) were laden with a 2mg/ml collagen gel layer with fluorescent beads (b). The
motions of fluorescent beads were tracked with a particle tracker program. Scale bar: 300pm
4.3.4.
Control over Contractile Directions
Cardiomyocytes contract along the sarcomere alignment which coincides with cell elongation.
Thus contractile direction can be controlled with surface patterning techniques, such as
microcontact printing (MCP) or nanopattern topography (NPT). Details of both methods can be
found in Sections 3.3.3 and 3.3.4 respectively. A schematic demonstration can be seen in Figure
4-3. In Chapter 3, surface patterning was used as a tool to induce cell alignment and study the
effect of alignment on the phenotypic changes of ESCDMs. In this chapter, we are able to use
the same tools to control contractile direction.
. 11:::.:
................
::::::::::::::
muu:::
::um-
I
Isotropy
*
0
Figure 4-3: Schematic Illustration of the Effect of Anisotropy on Contraction Direction.
The dotted arrows indicate the direction of contraction
4.4.
Results
4.4.1.
In-Phase Contraction is Detected across Monolayer
Two points were selected on the two ends of each cell and the changes in the distance between
the two points were recorded (Figure 4-1). We studied three cells separated by regions of noncontractile cells on the monolayer to eliminate the possibility that the changes in cell lengths
were attributed to simply being pulled by adjacent cells. By tracking the distance between the
two points, we demonstrated that the three cells shortened (contracted) and lengthened (relaxed)
in phase (Figure 4-4). This tracking method accurately recorded the rhythmic cycle of diastole
and systole of each cell. Similar to what is reported in the literature, cells spent two-thirds of the
time during each contraction cycle in diastole and rest in systole (160). In Figure 4-4, the small
amount of missing data is due to the low contrast of monolayer and the software was unable to
accurately detect the points of interest. The cell in the center (Cell 1) exhibited more vigorous
contraction resulting in a net change in length of close to 9% of the cell length, similar to the
maximum contraction strain observed for cardiomyocytes in vivo (155). For Cell 2 and Cell 3,
changes in cell lengths were smaller. This study demonstrated while individual contractions may
vary, a confluent monolayer of ESCDMs is able to contract in phase as long as they are
connected.
I'll,1. '--- :.....
.....
'-------l...................
-Cell
1
-Cell 2
a 1.06 -
--
1.04
8
Cell 3
1.02
1
0.98
,e
0.96
0.94 0.92-
0
1
3
2
4
5
6
Time (Sec)
Figure 4-4: In Phase Contraction of Cell Monolayer
Changes in cell length were measured with particle tracking. Three cells at three different regions
were tracked and they contract in unison. This result illustrates that ESCDMs cultured as a
confluent monolayer have the capability to contract in unison. The relative cell locations can be
found in Figure 4-1.
4.4.2.
Force Generated by Embryonic Stem Cell Derived Cardiomyocytes are
Comparable to That of Neonatal Myocytes
Forces generated by the ESCDM monolayer were estimated by tracking the movement of the
hydrogel, with a known elastic modulus, laden on top of the contracting monolayer. The
potential concern with gel-monolayer separation was mitigated since we have demonstrated that
ESCDM monolayer adhered strongly to the overlaying collagen gel after 11 days (Figure 3-7).
To estimate the movement, fluorescent beads were dispersed in the hydrogel during
polymerization and entrapped in the mesh network of the hydrogel. The movement of each bead
was monitored with particle tracking software (Figure 4-5). From inspecting movies of bead
87
..
....
....
....
contraction, we determined that the movement of beads was 3pm on average. For each
contraction, the movement of the beads was highly consistent (Figure 4-5 b&c).
(b)
Displacement over time
B
(c)
E1E
0)
06 2
4
L
8
10
12
14
16
18
'
'
Movement in x (pixels)
Time (sec)
Figure 4-5: Movement of a Fluorescent Bead
(a) Beads were monitored with particle tracking to determine the displacement due to ESCDM
contraction. During each contraction, 3pm of displacement was observed. Scale bar: 2pm. (b) the
magnitude of the displacement of the bead could be precisely measured and (c) the path line of the
bead was highly consistent.
To estimate the forces generated by the ESCDM monolayer from bead movements, a simplified
scaling analysis was deployed. The schematic illustration of the experimental setup can be seen
in Figure 4-6a where the beads close to the monolayer displace according to the contractions of
the monolayer. A simplified schematic is shown in Figure 4-6b and the following assumptions
88
are made: collagen gel displaces as a homogenous, isotropic and elastic (HIE) material with
Young's modulus (E), force per unit depth, P, is uniform and the collagen gel can be treated as a
semi-infinite domain. The assumption of collagen gel as an HIE material is valid because of the
small displacements of the beads. The assumption of treating collagen gel as a semi-infinite
domain is also legitimate since the thickness of the collagen gel is much larger than the depth at
which the contraction has an impact. The stress exerted by the monolayer can be described as
ca=E*,
where a is the normal stress and , is the strain.
a7 can be further scaled as
a~P/(Lc*Du)
where Lc is the characteristic length scale of the system of interest and Du is the unit depth.
, can be scaled as
c~2*S/Lc
The change in length is 2*6 assuming the beads displace in equal and opposite directions of 6
each.
Finally, by rearranging the equations, force per unit depth (P) can be scaled as
P- 2*E*6
The Young's modulus of 2mg/ml collagen I hydrogel is around lOKPa (146) and the
displacement is 3pm. Therefore, the force per unit depth out of the page is 60mN/m. For a
typical cluster with width of 100pm, we estimate that the force generated would therefore be
around 6pN. Another commonly used measure of forces generated by a monolayer is the stress
of the monolayer. Assuming the height of the monolayer is 5pm, the stress of the monolayer is
12mN/mm2 , which
agrees well with reported literature values for neonatal myocytes (53) (Table
2).
(a)
E,v
*
P
P
(b)
E,v
60
6
Figure 4-6: Schematic of Force Generation
Schematic of the setup to estimate the forces exerted by the ESCDM monolayer on the bead. This
illustration is upside down from the experimental setup since the collagen gel is laden on top of the
monolayer. In other words, the monolayer is sandwiched between the polystyrene culture substrate
and the hydrogel. P is the force per unit depth exerted by the ESCDM monolayer, 8 is the
displacement of the bead and E and v are the Young's modulus and kinematic viscosity of the
hydrogel.
4.4.3.
Direction of Contraction can be Controlled with Pattern Directions
For elongated cardiomyocytes, contractions occur along the long axis of the cell. Therefore, to
control the direction of the cell, one must control the direction of the elongation. As discussed in
the last chapter, both microcontact printing (MCP) and nanopattern topography (NPT) resulted in
elongated cells in specified directions (Figure 3-13). One example of the elongation and
contraction is shown in Figure 4-7. One isolated ESCDM was cultured on a NPT surface and
...
...........
.........
.........
I'll,
...........
"I'll"
................
.............
..
elongated along the direction of the grooves. When the cell contracted, it did so along the
principal direction of the cell which aligned well with the direction of the nanopatterns. The
change in length of the nuclear envelope, denoted with the blue double arrow lines, was on
average 26pm, 10% of the length of the nuclear envelope in diastole (Figure 4-7a). Similar
behavior was observed for cells seeded on MCP surface, with 9% net change in length between
contraction and relaxation (Figure 4-7b).
(a)
Relaxation
Contraction
Nanotopography
(b)
Microcontact
printing
0LIll
Figure 4-7: ESCDMs Contract along the Direction of Surface Patterning
ESCDMs are cultured on (a) the nanopatterned substrate and (b) microcontact printed surface. Cells
elongate and contract along the direction of the pattern (blue double arrow lines for NPT surface).
Scale bar: 50pm.
4.5. Discussions and Conclusions
Cardiomyocytes are contractile cells with the potential to assemble into miniature biomachines.
Embryonic stem cell derived cardiomyocytes, compared with neonatal myocytes, have the
advantage as an unlimited and continuous cell source. In this chapter, we demonstrated that
ESCDMs have equivalent capabilities as neonatal myocytes in becoming a source of
biomachines. Specifically, we conducted three key demonstrations in controlling ESCDMs for
biomachines applications. First, in phase contractions of ESCDM monolayer was demonstrated
by tracking changes in cell lengths for different cells on a confluent monolayer. Second, the
force generated by ESCDM monolayer was comparable to that by neonatal myocytes. Lastly, we
were able to control the contractile direction of cardiomyocytes by controlling the direction of
elongation.
Cardiomyocytes experience on average 10% strain during each contraction (155). In our
experiments, we observed strains smaller than 10%. This discrepancy between in vivo and in
vitro observations may be due to the difference in substrate stiffness, whether or not the adjacent
cells were contractile, and/or distributions of integrins. It has been shown that cardiomyocytes
exert maximal contractile forces on substrates with a similar stiffness as the myocardium which
is on the order of 50KPa, much softer than polystyrene (64,150). However, this theory does not
explain the heterogeneity in the degree of contraction. As observed in Figure 4-4, while some
ESCDMs contracted up to 9% strain, others contracted much less. It is possible that this is due to
a difference in integrin binding preventing well-adhered cells to elicit large displacements.
Alternatively, unlike the in vivo cell arrangements, some cardiomyocytes in vitro may be
surrounded by noncontractile cells and thus exhibit limited contractility.
Several methods to estimate cardiomyocyte forces have been developed and are illustrated in
Figure 4-8 and summarized in Table 2. Each one has its advantages and limitations. One method,
a variation of traction force microscopy involving seeding cardiomyocytes in an array of
micropillars and monitoring the movements of the pillars, has the difficulty in controlling how
cells adhere (Figure 4-8 a&b) (131). Another method directly probes the mechanics of single
cardiomyocytes by clamping them between two carbon fibers connected to a piezo-electric
translator probes or MEMS force transducer (Figure 4-8c,e&f) Single cell force measurements
ignore aggregate cell behavior which has been implicated to be key in force generation
(130,161,162). The direct contact between experimental probes and the cell is also likely to skew
the force measurements. The last technique measuring the bending of a thin PDMS film seeded
with cardiomyocytes assumes uniform bending and radius of curvature (Figure 4-8d) (53). The
reported force from their analysis (3 mN/mm2) agrees very well with our simplified scaling
analysis.
Our method aims to provide a preliminary estimate of the force measurements with a simple
scaling analysis and is able to produce force estimates which agree well with the literature.
Furthermore, our method does not involve the direct contact of the cardiomyocyte monolayer
with force measuring instrumentations, which can affect the results greatly depending on the
setup of individual experiments. This finding illustrates that ESCDMs can generate comparable
amount of forces as neonatal myocytes.
Neonatal myocytes
Micropillar bending
3.5pN
(single cells) (131)
Adult myocyte (single
cells) (130)
Unable to control cell
attachment
Piezoelectric translator
5.7pN
42 mN/mm2
Possible artifacts from clamping
Suspended single cells are not
..
........
- -
..........
60pN,
Neonatal myocytes
(monolayer) (53)
PDMS film bending
Adult myocyte (single
cells) (162)
MEMS force transducer
1-4mN/mm2
6pN,
l4mN/mm2
physiological relevant
Assuming ideal PDMS
curvature
Possible artifacts from clamping
Experimental artifact with glue
attachment
Suspended single cells are not
physiological
6pN,
12mN/mm2
Scaling analysis of bead
movements
Embryonic stem cell
derived myocytes
relevant
Simple scaling analysis
(monolayer)
Table 2: Force measurements of cardiomyocytes
Medium
Micropillar
(i
Halogen lamp
Cardiomyocyte
(d) PDMS Clamp Fixed
to Force Transducer
Muscular Thin Film
d
...
,z
PDMS Clamp Fixed
to Petri Dish
CCD
Figure 4-8: Different Methods of Measuring Forces of Cardiomyocytes
(a&b) Forces exerted by cardiomyocytes are inferred by the displacement of micropillars (131).
Neonatal myocytes exert forces around 3.5pN. (c) another method is to directly measure isolated
cardiomyocytes with carbon fibers and a piezo-electric translator. Adult cardiomyocytes exert 5.7pN
(130). (d) Feinburg et. al. cultured neonatal myocytes on a thin film of PDMS and measured the
deflection. Stress measurement is around 3 mN/mm 2 (53). (e&f) MEMS-based force transducer to
measure forces of single cardiomyocytes. Blue dashed lines represents the cell clamp area. (162)
Lastly, the direction of contractile cells can be controlled by surface patterning. This is essential
for future applications where the direction of actuation is important. Feinberg et. al. have used
aligned neonatal myocytes to produce linear and helical propellers (53). However, to our
knowledge, this is the first time surface patterning has been used to control the contractile
directions of ESCDMs.
By demonstrating in-phase contraction, estimating force generation and controlling contractile
directions, we illustrated the possibility to utilize ESCDMs as a possible continuous source of
contractile cells for miniature biomachines.
Chapter 5. Conclusions
5.1.
Significant Findings and Summary
The heart is the first functional organ during embryonic development and the last to cease
function upon death. Its primary role is to power the circulatory system by producing rhythmic
and controlled contractions. Its health is acutely linked to the overall health of an individual. To
break the heart down into its components, it is comprised of multiple cell types: cardiomyocytes
responsible for the contraction of the heart, cardiac fibroblasts producing appropriate
extracellular matrix proteins, endothelial cells and smooth muscle cells providing the vasculature
and neurons directing communications. Each cell type has its unique and essential role in the
health of the organ. The ECM network of the heart is equally dynamic and complex. Not only
does the network act as the supporting scaffold, but it also regulates cell organizations and
signaling. It has been indicated that various cardiovascular diseases (CVDs) result in misbalance
of collagen types and excess ECM remodeling (2,11,163,164).
Cardiovascular diseases are the leading cause of mortality with the total cost of the disease in
2010 estimated at $316.4 billion dollars in the United States (165). There is an urgent need for
more cost effective treatments for CVDs. Cell therapy has been considered as a possible
treatment option. However, many challenges remain, including but not limited to choosing the
right cell type, optimizing the delivery method, and integrating safely with the host tissue. The
benefit of cell therapy has not been unequivocally demonstrated suggesting the need for an
improved mechanistic understanding.
In vitro cell culture platforms are one solution to simplify and systematically characterize the
effects of individual biochemical or biophysical factors in a complex biological phenomenon. In
this thesis, in vitro tissue engineering techniques were utilized to characterize the differentiation
and subsequent maturation processes of cardiomyocytes derived from murine embryonic stem
cells (ESCs). It has been reported that ESC derived cardiomyocytes (ESCDMs) exhibit
phenotypes and characteristics similar to fetal cardiomyocytes, rather than adult cardiomyocytes.
The mechanism by which cells progress to exhibit more mature phenotypes is unclear and
challenging to elucidate in vivo. Furthermore, how differentiation is affected by the densely
packed and highly dynamic environment of the heart also remains to be answered. We addressed
these challenges with in vitro culture systems.
First, to study differentiation in vitro, two biophysical cues, culture dimension and uniaxial
cyclic stretch, were investigated by developing a microfluidic platform that mimics the dynamic
environment of the heart. A confined environment, such as a microfluidic channel, was found to
promote differentiation. This is partially due to the accumulation of a cell-secreted molecule,
bone morphogenetic protein 2, which has been implicated as cardiogenic. Second, uniaxial cyclic
stretch is a negative regulator of differentiation. The negative regulation can be traced back to
developmental biology where contractions do not occur until 22 days after conception for
humans (57). Thus, it is unlikely for cells to experience uniaxial strain during differentiation in
vivo. This finding has implications on the timing of introducing stem cells into the mechanicallyactive myocardium.
ESCDMs were further isolated upon differentiation in order to conduct specific studies on how
they continue to mature in vitro. Again, two biophysical factors, motivated by the natural
biophysical environment of the heart, were studied - cell alignment and the presence of collagen
I, the most abundant ECM protein in the adult myocardium. Three assessment measurements
were deployed to quantify the phenotypic progressions of ESCDMs: cell shape change
98
(elongation to become more rod-shaped), sarcomere development (organization with specific
periodicity) and binucleation (higher percentage in mature cardiomyocytes).
Cell alignment was induced with microcontact printing (MCP) of 20/20pm lines or nanopattem
topography (NPT) with 700nm/3.5pm grooves and ridges. While both techniques resulted in cell
elongation, cells on MCP surfaces reached maximum elongation within 48 hours and did not
increase over time and cells on NPT surfaces continued to elongate over time up to two weeks.
Sarcomeres were well-developed on both surfaces and the width of sarcomeres was highly
dependent on the periodicity of the NPT surface. In terms of binucleation, the MCP surface alone,
without the addition of collagen, slightly but significantly increased the percentage of
binucleation compared with a smooth surface. In contrast, compared with the isotropic case,
ESCDMs cultured on NPT surface observed markedly reduced binucleation, due to increased
cytokinesis. This is the first report on the effects of surface patterning on binucleation of
ESCDMs.
To study the effect of collagen I on phenotypic changes of ESCDMs, collagen I was presented to
the cells in three forms - cells were mixed into collagen hydrogel during polymerization,
collagen hydrogel was laden on top of a confluent monolayer or exogenous soluble collagen was
added to the medium. When cells were suspended in collagen gel, they did not elongate, extend
or develop sarcomere structures. ESCDM monolayer laden with collagen gel exhibited a reduced
elongation suggesting that the gel may impede cell motility and shape change. Lastly, soluble
collagen markedly reduced binucleation by promoting cytokinesis. This suggests that the effect
of collagen I on ESCDMs depends on how it is presented and soluble collagen may play an
important role in promoting ESCDM proliferation.
In addition to influencing differentiation and maturation by imposing biophysical stimulations,
we further explored the possibility to extend our control of ESCDMs for non-medical
applications. In order to do so, three essential proof-of-concept demonstrations were developed
to illustrate that ESCDMs are capable of contracting in phase along a specified direction and the
amount of force generated is comparable to that of neonatal myocytes, the primary choice of
myocyte-powered devices. In many myocyte-powered microdevices, the intrinsic abilities for
cells to assemble and align with proper cues are often overlooked. Here, we were able to use the
tools developed and lessons learned in studying differentiation and maturation of ESCDMs in the
application into biomachines.
In this thesis work, we were able to systematically explore the differentiation and maturation of
ESCDMs with various biophysical factors. The specific findings on how particular biophysical
stimulations affect differentiation and maturation open up new avenues of exploration.
Furthermore, we were able to take undifferentiated embryonic stem cells all the way to purified
cardiomyocytes with maturing phenotypes and demonstrate the potential to use them as
miniature actuators. This thesis lays the foundation to utilize embryonic stem cells to
differentiate into cardiomyocytes and other cell types, either for mechanistic studies or for nonmedical applications.
100
5.2.
Are We There Yet?
One main motivation behind this thesis is to characterize the cardiogenesis process, from
undifferentiated embryonic stem cells to adult cardiomyocytes, in order to replenish infracted
heart tissue with functional ESCDMs. As described in the last section, ESCDM differentiation
and maturation can be enhanced with external biophysical and biochemical stimuli, they
however still fall short of exhibiting adult myocyte phenotypes. Figure 5-2 presents a simplified
time progression of the phenotypic signatures documented in this thesis and the corresponding
changes in vivo. Although directly comparing in vitro results with in vivo ones might be
speculative, it provides a general direction of how in vitro findings can be related back to a
physiologically relevant condition. For instance, percentage of binucleation increases both in
vivo and in vitro despite that the trends do vary. The maximum binucleation observed in our
study is 20% (Figure 3-15) which corresponds to 4-day old myocytes, more mature than 1-2 day
neonatal myocytes which are the gold standard for most experiments. This however, is far short
of the 90% of binucleated cells observed for adult myocytes. Additional stimuli and longer
culture time are likely to be necessary to further enhance maturation (Also see section 5.3.3 for
more discussions).
Furthermore, as observed in mature cardiomyocytes, ESCDMs cultured up to 22 days in this
thesis exhibited clear and organized sarcomeres and displayed a rod-like shape (Figure 5-1)(166).
Cardiomyocytes elongate during development, from an aspect ratio (AR) of 1 for fetal myocytes
to 4 for neonatal myocytes and eventually 7 for adult myocytes. ESCDMs also elongated in vitro
to an AR of 3.5 on isotropic surface and 6 on nanopatterned surface. Taken together, with
appropriate biophysical stimuli, we were able to start from undifferentiated embryonic stem cells
and differentiate them to exhibit phenotypes slightly more mature than neonatal myocytes.
101
: :..........
111 "1 '..........
2 Days ESCDM
21 Days ESCDM
Freshly Harvested
Adult Myocyte
Figure 5-1: Sarcomere organization of ESCDMs compared with adult cardiomyocytes
Sarcomere organization was visualized with immunofluorescent staining of sarcomere a-actinin.
ESCDMs shown here were cultured on isotropic surface. Much clear sarcomere structures are
observed for 21 day ESCDMs compared with 2 day ESCDMs. The sarcomere organization of 21 day
ESCDMs resembles the freshly harvested adult cardiomyocytes (166). Scale bar: 10pm
In this study, we have demonstrated that ESCDMs can be a replacement for neonatal myocytes
and have the advantage of eliminating fresh cell harvests and the possible variability due to
animal differences. However, there are some obstacles preventing ESCDMs to be widely used as
a replacement of neonatal myocytes in vitro and cell therapy in vivo.
The greatest challenge is scalability. In embryoid body based assays, only 2-8% of all cells
differentiated from embryonic stem cells are cardiomyocytes. In order to have a large amount of
cardiomyocytes, cost-prohibitively large quantities of media and reagents are required. For
instance, to replace the 1 billion necrotic cardiomyocytes lost during an average MI with
ESCDMs, a minimum of 12.5 billion cells have to be differentiated requiring $2.5 million for
reagents and media without considering the labor cost. Recent advancements in varying
exogenous growth factors over time or co-culturing with endoderm-like cells in a serum-free
102
medium have resulted in higher yields of 10-25% cardiomyocytes (46,109,167). How to costeffectively and efficiently produce large quantity of pure ESCDMs for therapeutic applications
remains a major hurdle to overcome (168).
Other challenges yet to be overcome include further maturation of ESCDMs, inhibition of
teratoma formation in vivo (18), and survival and integration of ESCDMs to the host tissue (49).
We have come a long way in characterizing the cardiogenesis process, identifying additional
mechanisms and stimuli responsible for differentiation, but much more research is required to
solve the aforementioned problems in order to truly harness the power of stem cell-based therapy.
103
.......................................
Protein
Marker
Day 0
Undifferentiated
ESCs
Day 1-3
Mesoderm
Day 4-7
Early Differentiation
Committed to cardiac
lineage
-Oct
hEn
cm
N
NKx 2.5
a)
g
Sarcomnere
C6
r
a-MHC
Day 8 -
Late Differentiation
Contraction observed
1-2 days post birth
Neonatal myocytes
(gold standard)
4
racllyury
-
Experimental Observations In vivo and In vitro
Procedures
FceH shiape
L chane
Bincleation
q
~0
MLC-2
-
eveopment
C.-
M>
E~E
~~Fluorescent
ESCDMs
O
C5,1
T- -, r
0f
aO
.
"e
c0
.tmE
W
Ea)U'
E
oEC
Z
0C
O-
W U) V
~C,4
0~
cI
Con~n(
2 ET--
04
CL~~~C
C
~~~~U
uu M ~'
.E
_
.0.R/
8 week +
Adult myocytes
Figure 5-2: Time course of phenotypic changes of ESCDMs
This time course illustration combines results found in this thesis with existing literature. The
information on protein markers are adapted from Boheler et. al. (69). 1: binucleation data in vivo
were adapted from Soonpaa et. al. (125). 2: quantification and details can be found in Figure 3-9. The
numbers are for ESCDMs cultured on isotropic surface without the addition of collagen I. The
amount of binucleation agrees well with existing in vitro data (126). 3: the description of
sarcomerogenesis was adapted from Sparrow et. al. (127). 4: the images where the description was
based from can be seen in Figure 3-6. 5: Aspect ratio of cardiomyocytes at different developmental
stage was extracted from Seki et. al. (133). 6: Aspect ratio of ESCDMs can be seen in Figure 3-13.
Abbreviations: EM - embryonic cardiomyocytes, FM - fetal myocytes, NM - neonatal myocytes, AM
- adult myocytes, AR - aspect ratio
104
5.3.
Future Directions
This thesis work has generated interesting findings and results, but more importantly, it serves as
a spring board for further research. Here are some of my suggestions of the future directions of
this research.
5.3.1.
Microfluidic Platform Development
First, in this study, microfluidic devices were found to be a superior platform, compared with
conventional well plates, in studying differentiation presumably because of the restricted
transport. In a solid organ, such as the heart, cells are tightly packed and surrounded by
extracellular matrix. The transport phenomenon is likely to be more restricted than that in a
conventional well-plate system. The advantages of microfluidic platform can potentially
challenge the conventional cell culture methods and increase the sensitivity of any assay. Thus,
more effort should be directed toward improving the microfluidic platform to mimic the in vivo
environment by including a 3D extracellular matrix, other cell types and appropriate stimuli.
Such improved microfluidic platform can be used to study not only the differentiation and
maturation of ESCDMs, but also other complex ensemble cell processes.
5.3.2.
Further Investigation into Binucleation
In terms of maturation, the finding of binucleation reduction with nanopatterns and soluble
collagen are highly intriguing for multiple reasons. While we have demonstrated that the
decrease in binucleation is actually a result of an increase in cytokinesis, we have yet to explore
the signaling pathways responsible for this finding for cells on an NPT surface or exposed to
soluble collagen I. For the studies with NPT, a variation of NPT surfaces can be tested to see if
the increase of cytokinesis is universal for all periodicity. Analogously, a titration study can be
performed to investigate if the decrease in binucleation is dose-dependent on collagen I
105
concentration. Furthermore, mechanistic biochemical assays can be performed to gain a better
understanding of the underlying signaling pathways. Little, if any, is known about how nanoscale
topography is translated into biochemical signals or how soluble collagen I signals differently
than collagen I hydrogel in the context of ESCDMs maturation. As a result, techniques targeting
several proteins, such as RT-PCR, antibody neutralization, knock-in/out methods, or
immunofluorescent staining, are unlikely to be effective since it is not clear which proteins
should be investigated. A large scale screening technique, such as microarray, may be necessary
to look for activated proteins. This finding can have profound implications on understanding
cardiomyocyte proliferation.
Another avenue of investigation is to test whether or not this finding can be reproduced in vivo.
The reason or the benefit for higher percentage of binucleated cells in adult myocytes is unclear.
There is one school of thought that multinucleated cells are cells unable to finish the cell cycle
and are frozen in a suboptimal state. If this is true, then one can supplemental soluble collagen I
or design implantable biomaterials with nanoscale topography to promote cytokinesis.
Eventually, in vivo studies can be performed to see if soluble collagen I or nanoscale topography
can improve the ejection fraction of an injured heart in animal models.
5.3.3.
Incorporation of Additional Stimulations
In this thesis, we investigated several biophysical and biochemical stimuli - uniaxial cyclic
stretch, culture dimension, interaction with collagen I and surface patterning. However, there are
other cues that may provide further insights on the cardiogenesis process. For instance, electrical
stimulation has been demonstrated to induce electrical coupling of ESCDM (60,156). Thus, it
will be interesting to examine the combinatorial effect of electrical and mechanical stimuli on
ESCDMs.
106
Even within the study of a single stimulation, there are multiple variables in the experimental
design, including the duration of stimulation application, magnitude and frequency of application
and output parameters. Consistent and systematic variations of parameters and designs of
experiments should be conducted across research groups so results can be reliably compared.
Parametric studies should be performed to determine the relative sensitivities of the outputs to
input parameter variation. For instance, by compiling existing literature studying effect of
mechanical stretch on differentiation, it appears that stretch duration or stretch direction
(equibiaxial or uniaxial) may be less significant than strain rate (Table 3). In the concerted effort
to characterize and understand how differentiation can be augmented biophysically and
biochemically, standardized protocols should be adapted.
hESCs
1.67 %
Equibiaxial
Day 0 - Day 12
Inhibition(107)
mESCs
10%
Uniaxial (in gel)
Day 24 - Day 27
Inhibition(81)
mESCs
30%
Uniaxial (in gel)
Day 24 - Day 27
Promotion(81)
mESCs
Static
10-20% strain
Equibiaxial
2 hours on Day 4
Promotion(79)
Table 3: Effect of mechanical stretch on differentiation
5.3.4.
Investigation of Maturation of ESCDM Electrical Properties
The maturation of the electrophysiological properties of ESCDMS, such as conduction velocity
and ion channel development, was outside of the scope of this thesis but are critical during
107
development (69,116,133,169). Compared with terminally developed cardiomyocytes, immature
ESCDMs have smaller action potentials and exhibit more atrial-like electrical signature. For
future studies, incorporating the electrophysiological properties of ESCDMs as a maturation
output can provide a more comprehensive picture of the maturing progress.
5.3.5.
Incorporation of Patterning in 3D and Other Cell Types
The surface patterns used in this study is limited to two dimensions. However, the environment
of the heart is highly three-dimensional. It will be interesting to incorporate patterning techniques
into three-dimensional matrix. Aligned collagen I has been developed by stretching collagen
during polymerization or imposing shear flow on collagen fibers in microfluidic channels
(170,171). Collagen hydrogels used in in vitro experiments usually have low stiffnesses
compared with the stiffness of the native myocardium. Thus, it is worthwhile to develop
techniques to produce stiffer collagen gels with aligned collagen fibers. This will be a more
realistic condition of the heart.
As mentioned in Chapter 1, cardiomyocytes only account for 35% of the cells in terms of total
cell numbers. The other 65% include cardiac fibroblasts, endothelial cells and neurons. Therefore,
it is only reasonable to eventually incorporate other cell types into the in vitro system to study
cell-cell interactions and behaviors. Current studies suggest that non-myocytes may be critical in
providing the proper formation of the tissue construct although systematic characterizations are
still lacking (55,172).
5.3.6.
Self Assembly
In this thesis work, we precisely controlled the biophysical environment with techniques such as
uniaxial stretch and surface patterning to direct cell behavior. There can be a different approach
108
which relies on the ability for cells to self-assemble with minimum cues. Studies have shown that
fibroblasts are able to self assemble in a non-adherent mold (173,174). It is foreseeable that
ESCDMs are also able to self assemble. This development has the potential in building cellular
units as part of the biomachines.
5.3.7.
Concluding Thoughts
Winston Churchill once said "Out of intense complexities intense simplicities emerge." In this
thesis, I have portrayed a rather complex picture of the differentiation and maturation processes
of ESCDMs. This work contributes to the larger body of research dedicated to understanding
cardiogenesis. It is my hope that a simple yet consistent framework will emerge from the
plethora of data and results and this framework can help shape the way cell therapy can be used
to treat cardiovascular diseases and we can obtain a better understanding of population based cell
behavior.
109
Appendix I - Embryonic Stem Cell Maintenance and Differentiation Protocol
Embryonic Stem Cell Medium Formulation (100ml): ESCM
Amount
79.3ml (if
Invitrogen)
Source and Cat#
Invitrogen (G-MEM, BHK21)
iml
100pl
Sigma (M7145)
Sigma (M7522)
iml
Sigma (S8636)
Penicillin-Streptomycin
HEPES
Iml
2.5ml
Sigma (P0781)
Sigma (H4034) Powder
Knockout Serum
Replacement
Leukemia Inhibitory
Factor
15ml
Invitrogen (10828-028)
100plI
Sigma (L5158)
iml
Sigma (G7513)
GMEM (Glasgow
Minimum Essential
Medium)
100x NEAA
2-Mercaptoethanol (2ME)
Sodium Pyruvate
Detail
Stock concentration: 10-1M
Need: 10-4M in solution
Stock concentration: 100mM
Need: 1mM
100x stock
Stock Concentration: IM
Want 25mM
15% in final concentration
Final concentration: 1000U/ml
Sigma: Stock Concentration
10pg/ml (no less than
108U/mg-> thus, need total of
1pVg for medium
L-Glutamine
+
If using Sigma GMEM, need to
add. Stock 200mM solution.
Final concentration: 2mM
If using Invitrogen GMEM, no
need
Usually add everything and filter
0.1% Gelatin Solution
20mg of gelatin (cherry's black cabinet space) + 20ml 1x PBS-/- -> Wet Autoclave
The sterile solution should be refrigerated
2x Freezing Solution (20%DMSO in ESCM)
2ml of DMSO + 8ml of ESCM
1M HEPES Solution
23.8g of HEPES powder in 100ml of Millipore Water -> filter
10-'M 2-ME Solution
110
100pl)
1Opl of Stock (14.3M) + 1.42ml 1xPBS-/- - 0.2pm filter (good for one month)
10-IM Ascorbic Acid
0.99g of ascorbic acid + 5ml of sterile water -> 0.2pm filter (goof for one month)
Differentiation Medium
Replace KOSR with ESC-FBS (Invitrogen: 16141079) & remove LIF
Procedure
Coat Flasks
0.1% Gelatin for 30min in incubator; aspirate
Cell Passage (every other day usually. Important not to have it more than 75% confluent)
Aspirate medium and wash cells with PBS twice
Add 0.5ml of trypsin-EDTA (0.05%)
Incubate for a few minutes
Add 5ml of ESCM
Count cells
Spin (1500rpm, 5min)
Resuspend in ESCM
300,000 cells/T25 (by the time, typically cell numbers increase 5 times in 2 days)
6-6.5ml of media/T25
Thaw
Transfer to 37C water bath
Transfer to centrifuge tube
Add 5ml ESCM gradually
Spin (1500rpm, 5min)
Suspend in 6ml ESCM in T25
Change medium the next day
Freeze
Suspend 2-3x106cells/0.25ml of ESCM
Add equal volume of freeze solution and cell solution
Transfer cells to freeze vial: 0.5ml/vial
Keep in -80C overnight
Long-term storage: keep them in the gas phase of N2 tank (Shelf: A7)
Hanging Drop
111
Suspend cells in 3-4x10 6 cells/ml of ESCM
Take 50,000 cells for 5ml of Differentiating medium (DM) (Make cell concentration of
10,000cell/ml)
Place 10-15ml sterile 1x PBS in 10cm petri dish
Invert the cover and place 3 0 pl droplet of cell suspension solution on the cover
(Wednesday)
Place in incubator for 2 days
Collect the droplets and place in nonadherent plastic dish (not tissue culture treated) with
15ml of DM (Friday)
EBs ready for use on Monday
Dissociation of Embryoid Bodies
Wash cells with 1x PBS -/- twice or trypsin once
Add 0.05% trypsin for 5-7min (enough to cover -0.5-lml), pipette gently to dislodge EBs
from plastic surface after 5-7 min
Add 2ml medium to each well and transfer all into a falcon tube
Centrifuge (1,200rpm, 5min)
Remove supernatant and add -3ml trypsin into tube and incubate for 5-7min
Every 2-3 minutes, agitate the tube gently
Pipette slowly and observe small amount to see if they have become single cell
suspensions
Add equal amount of medium and centrifuge (1,200rpm, 5min)
Remove supernatant and resuspend in media
112
.............
- -.:.::::
::-.........
........
..............
.......
Appendix II - Cardiomyocyte Enrichment with Percoll Separation
Objective - Enrich ESCDM population
Procedure
Put 3ml of 40.5% Percoll in one 15ml tube
Use a 2ml pipet, VERY gently underlay 58.5% layer on the bottom of the same tube. (should
take 3-5 minutes to add 3ml in; should see a line of separation)
Gently add cell solution on top (now should see 3 layers)
Centrifuge for 1,500g for 30 min
After centrifugation, should see cell layers between medium and 40.5% Percoll, between two
layers of Percoll, and a pellet at the bottom.
Gently remove acellular medium and first layer of Percoll
Collect the cell layer between two percoll densities in a 15ml falcon tube and add 5ml media
add 2ml of media to the bottom layer of Percoll
Centrifuge the two tubes (1,200rpm, 5min)
Resuspend and plate
(Enriched cardiomyocyte layers denoted in Blue)
Before
After
centrifugation centrifugation
3ml of 40.5% Percoll Solution
Percoll: 1.215 ml
IM NaCl: 0.45ml (to make 150mM final
concentration)
IM HEPES: 0.06ml (to make 20mM final
concentration)
Water: 1.275m1
3ml of 58.5% Percoll Solution
Percoll: 1.755 ml
IM NaCl: 0.45ml (to make 150mM final
concentration)
IM HEPES: 0.06ml (to make 20mM final
concentration)
Water: 0.735 ml
113
References
(1) Rosamond W, Flegal K, Friday G, Furie K, Go A, Greenlund K, et al. Heart disease and
stroke statistics - 2007 update - A report from the American Heart Association Statistics
Committee and Stroke Statistics Subcommittee. Circulation 2007 FEB 6;115(5):E69-E171.
(2) Weber KT. Cardiac interstitium in health and disease: the fibrillar collagen network
J.Am.Coll.Cardiol. 1989 Jun; 13(7):1637-1652.
(3) Borg TK, Caulfield JB. The collagen matrix of the heart Fed.Proc. 1981 May 15;40(7):20372041.
(4) Borg TK, Ranson WF, Moslehy FA, Caulfield JB. Structural basis of ventricular stiffness
Lab.Invest. 1981 Jan;44(1):49-54.
(5) LeGrice IJ, Smaill BH, Chai LZ, Edgar SG, Gavin JB, Hunter PJ. Laminar structure of the
heart: ventricular myocyte arrangement and connective tissue architecture in the dog
Am.J.Physiol. 1995 Aug;269(2 Pt 2):H571-82.
(6) Helm P, Beg MF, Miller MI, Winslow RL. Measuring and mapping cardiac fiber and laminar
architecture using diffusion tensor MR imaging Ann.N.Y.Acad.Sci. 2005 Jun; 1047:296-307.
(7) Wu Y, Wu EX. MR investigation of the coupling between myocardial fiber architecture and
cardiac contraction Conf.Proc.IEEE Eng.Med.Biol.Soc. 2009;2009:4395-4398.
(8) Beyar R, Yin FC, Hausknecht M, Weisfeldt ML, Kass DA. Dependence of left ventricular
twist-radial shortening relations on cardiac cycle phase Am.J.Physiol. 1989 Oct;257(4 Pt
2):H1119-26.
(9) Espira L, Czubryt MP. Emerging concepts in cardiac matrix biology
Can.J.Physiol.Pharmacol. 2009 Dec;87(12):996-1008.
(10) Little CD, Rongish BJ. The extracellular matrix during heart development Experientia 1995
Sep 29;51(9-10):873-882.
(11) Burgess ML, McCrea JC, Hedrick HL. Age-associated changes in cardiac matrix and
integrins Mech.Ageing Dev. 2001 Oct; 122(15):1739-1756.
(12) Ross RS, Borg TK. Integrins and the myocardium Circ.Res. 2001 Jun 8;88(11):1112-1119.
(13) Pope AJ, Sands GB, Smaill BH, LeGrice IJ. Three-dimensional transmural organization of
perimysial collagen in the heart Am.J.Physiol.Heart Circ.Physiol. 2008 Sep;295(3):H1243H1252.
114
(14) MacKenna DA, Omens JH, McCulloch AD, Covell JW. Contribution of collagen matrix to
passive left ventricular mechanics in isolated rat hearts Am.J.Physiol. 1994 Mar;266(3 Pt
2):H1007-18.
(15) Fomovsky GM, Thomopoulos S, Holmes JW. Contribution of extracellular matrix to the
mechanical properties of the heart J.Mol.Cell.Cardiol. 2010 Mar;48(3):490-496.
(16) Ott HC, Matthiesen TS, Goh SK, Black LD, Kren SM, Netoff TI, et al. Perfusiondecellularized matrix: using nature's platform to engineer a bioartificial heart Nat.Med. 2008
Feb;14(2):213-221.
(17) Johnson P, Maxwell DJ, Tynan MJ, Allan LD. Intracardiac pressures in the human fetus
Heart 2000 Jul;84(1):59-63.
(18) Laflamme MA, Murry CE. Regenerating the heart. Nat.Biotechnol. 2005 JUL;23(7):845856.
(19) Gerecht-Nir S, Radisic M, Park H, Cannizzaro C, Boublik J, Langer R, et al. Biophysical
regulation during cardiac development and application to tissue engineering. Int.J.Dev.Biol.
2006;50(2-3):233-243.
(20) Shimizu T, Yamato M, Isoi Y, Akutsu T, Setomaru T, Abe K, et al. Fabrication of pulsatile
cardiac tissue grafts using a novel 3-dimensional cell sheet manipulation technique and
temperature-responsive cell culture surfaces. Circ.Res. 2002 FEB 22;90(3):E40-E48.
(21) Zimmermann WH, Melnychenko I, Eschenhagen T. Engineered heart tissue for regeneration
of diseased hearts Biomaterials 2004 Apr;25(9):1639-1647.
(22) Zimmermann WH, Schneiderbanger K, Schubert P, Didie M, Munzel F, Heubach JF, et al.
Tissue engineering of a differentiated cardiac muscle construct. Circ.Res. 2002 FEB 8;90(2):223230.
(23) Anversa P, Nadal-Ginard B. Myocyte renewal and ventricular remodelling. Nature 2002
JAN 10;415(6868):240-243.
(24) Taylor DA, Atkins BZ, Hungspreugs P, Jones TR, Reedy MC, Hutcheson KA, et al.
Regenerating functional myocardium: improved performance after skeletal myoblast
transplantation Nat.Med. 1998 Aug;4(8):929-933.
(25) Menasche P, Alfieri 0, Janssens S, McKenna W, Reichenspurner H, Trinquart L, et al. The
Myoblast Autologous Grafting in Ischemic Cardiomyopathy (MAGIC) trial: first randomized
placebo-controlled study of myoblast transplantation Circulation 2008 Mar 4; 117(9):1189-1200.
(26) Reinecke H, Poppa V, Murry CE. Skeletal muscle stem cells do not transdifferentiate into
cardiomyocytes after cardiac grafting J.Mol.Cell.Cardiol. 2002 Feb;34(2):241-249.
115
(27) Rubart M, Soonpaa MH, Nakajima H, Field U. Spontaneous and evoked intracellular
calcium transients in donor-derived myocytes following intracardiac myoblast transplantation
J.Clin.Invest. 2004 Sep; 114(6):775-783.
(28) Jackson KA, Majka SM, Wang H, Pocius J, Hartley CJ, Majesky MW, et al. Regeneration
of ischemic cardiac muscle and vascular endothelium by adult stem cells J.Clin.Invest. 2001
Jun;107(11):1395-1402.
(29) Orlic D, Kajstura J, Chimenti S, Bodine DM, Leri A, Anversa P. Bone marrow stem cells
regenerate infarcted myocardium Pediatr.Transplant. 2003;7 Suppl 3:86-88.
(30) Quaini F, Urbanek K, Beltrami AP, Finato N, Beltrami CA, Nadal-Ginard B, et al.
Chimerism of the transplanted heart N.Engl.J.Med. 2002 Jan 3;346(1):5-15.
(31) Alvarez-Dolado M, Pardal R, Garcia-Verdugo JM, Fike JR, Lee HO, Pfeffer K, et al. Fusion
of bone-marrow-derived cells with Purkinje neurons, cardiomyocytes and hepatocytes Nature
2003 Oct 30;425(6961):968-973.
(32) Murry CE, Soonpaa MH, Reinecke H, Nakajima H, Nakajima HO, Rubart M, et al.
Haematopoietic stem cells do not transdifferentiate into cardiac myocytes in myocardial infarcts
Nature 2004 Apr 8;428(6983):664-668.
(33) Nygren JM, Jovinge S, Breitbach M, Sawen P, Roll W, Hescheler J, et al. Bone marrowderived hematopoietic cells generate cardiomyocytes at a low frequency through cell fusion, but
not transdifferentiation Nat.Med. 2004 May; 10(5):494-50 1.
(34) Gnecchi M, He H, Liang OD, Melo LG, Morello F, Mu H, et al. Paracrine action accounts
for marked protection of ischemic heart by Akt-modified mesenchymal stem cells Nat.Med.
2005 Apr; 11(4):367-368.
(35) Kinnaird T, Stabile E, Burnett MS, Shou M, Lee CW, Barr S, et al. Local delivery of
marrow-derived stromal cells augments collateral perfusion through paracrine mechanisms
Circulation 2004 Mar 30;109(12):1543-1549.
(36) Uemura R, Xu M, Ahmad N, Ashraf M. Bone marrow stem cells prevent left ventricular
remodeling of ischemic heart through paracrine signaling Circ.Res. 2006 Jun 9;98(11):14141421.
(37) Hansson EM, Lindsay ME, Chien KR. Regeneration next: toward heart stem cell
therapeutics Cell.Stem Cell. 2009 Oct 2;5(4):364-377.
(38) Agbulut 0, Mazo M, Bressolle C, Gutierrez M, Azarnoush K, Sabbah L, et al. Can bone
marrow-derived multipotent adult progenitor cells regenerate infarcted myocardium?
Cardiovasc.Res. 2006 Oct 1;72(1):175-183.
116
(39) Braun T, Martire A. Cardiac stem cells: paradigm shift or broken promise? A view from
developmental biology. Trends Biotechnol. 2007 OCT;25(10):441-447.
(40) Anversa P, Kajstura J, Leri A, Bolli R. Life and death of cardiac stem cells - A paradigm
shift in cardiac biology. Circulation 2006 MAR 21;113(11):1451-1463.
(41) Barbash IM, Chouraqui P, Baron J, Feinberg MS, Etzion S, Tessone A, et al. Systemic
Delivery of Bone Marrow-Derived Mesenchymal Stem Cells to the Infarcted Myocardium:
Feasibility, Cell Migration, and Body Distribution. Circulation 2003 August 19;108(7):863-868.
(42) Xaymardan M, Tang L, Zagreda L, Pallante B, Zheng J, Chazen JL, et al. Platelet-Derived
Growth Factor-AB Promotes the Generation of Adult Bone Marrow-Derived Cardiac Myocytes.
Circ Res 2004 March 19;94(5):e39-45.
(43) Deb A, Wang S, Skelding KA, Miller D, Simper D, Caplice NM. Bone Marrow-Derived
Cardiomyocytes Are Present in Adult Human Heart: A Study of Gender-Mismatched Bone
Marrow Transplantation Patients. Circulation 2003 March 11;107(9):1247-1249.
(44) Di Nardo P, Forte G, Ahluwalia A, Minieri M. Cardiac progenitor cells: potency and control
J.Cell.Physiol. 2010 Sep;224(3):590-600.
(45) Kehat I, Kenyagin-Karsenti D, Snir M, Segev H, Amit M, Gepstein A, et al. Human
embryonic stem cells can differentiate into myocytes with structural and functional properties of
cardiomyocytes J.Clin.Invest. 2001 Aug; 108(3):407-414.
(46) Mummery CL, Ward D, Passier R. Differentiation of human embryonic stem cells to
cardiomyocytes by coculture with endoderm in serum-free medium Curr.Protoc.Stem Cell.Biol.
2007 Jul;Chapter 1:Unit 1F.2.
(47) Laflamme MA, Chen KY, Naumova AV, Muskheli V, Fugate JA, Dupras SK, et al.
Cardiomyocytes derived from human embryonic stem cells in pro-survival factors enhance
function of infarcted rat hearts Nat.Biotechnol. 2007 Sep;25(9):1015-1024.
(48) van Laake LW, Passier R, Monshouwer-Kloots J, Verkleij AJ, Lips DJ, Freund C, et al.
Human embryonic stem cell-derived cardiomyocytes survive and mature in the mouse heart and
transiently improve function after myocardial infarction Stem Cell.Res. 2007 Oct; 1(1):9-24.
(49) Segers VFM, Lee RT. Stem-cell therapy for cardiac disease. Nature 2008 FEB
21;451(7181):937-942.
(50) Tanaka Y, Sato K, Shimizu T, Yamato M, Okano T, Kitamori T. A micro-spherical heart
pump powered by cultured cardiomyocytes Lab.Chip 2007 Feb;7(2):207-212.
(51) Tanaka Y, Morishima K, Shimizu T, Kikuchi A, Yamato M, Okano T, et al. An actuated
pump on-chip powered by cultured cardiomyocytes Lab.Chip 2006 Mar;6(3):362-368.
117
(52) Kim J, Park J, Yang S, Baek J, Kim B, Lee SH, et al. Establishment of a fabrication method
for a long-term actuated hybrid cell robot Lab.Chip 2007 Nov;7(11):1504-1508.
(53) Feinberg AW, Feigel A, Shevkoplyas SS, Sheehy S, Whitesides GM, Parker KK. Muscular
thin films for building actuators and powering devices Science 2007 Sep 7;317(5843):1366-1370.
(54) Xi J, Schmidt JJ, Montemagno CD. Self-assembled microdevices driven by muscle Nature
Materials 2005; 2005;4(2):180 <last-page> 184.
(55) Zimmermann WH, Fink C, Kralisch D, Remmers U, Weil J, Eschenhagen T. Threedimensional engineered heart tissue from neonatal rat cardiac myocytes Biotechnol.Bioeng. 2000
Apr 5;68(1):106-114.
(56) Guo XM, Zhao YS, Chang HX, Wang CY, E LL, Zhang XA, et al. Creation of engineered
cardiac tissue in vitro from mouse embryonic stem cells Circulation 2006 May 9; 113(18):22292237.
(57) Stock UA, Vacanti JP. Cardiovascular physiology during fetal development and
implications for tissue engineering Tissue Eng. 2001 Feb;7(1):1-7.
(58) Singhvi R, Kumar A, Lopez GP, Stephanopoulos GN, Wang DI, Whitesides GM, et al.
Engineering cell shape and function Science 1994 Apr 29;264(5159):696-698.
(59) Bursac N, Parker KK, Iravanian S, Tung L. Cardiomyocyte cultures with controlled
macroscopic anisotropy: a model for functional electrophysiological studies of cardiac muscle
Circ.Res. 2002 Dec 13;91(12):e45-54.
(60) Au HTH, Cui B, Chu ZE, Veres T, Radisic M. Cell culture chips for simultaneous
application of topographical and electrical cues enhance phenotype of cardiomyocytes. Lab Chip
2009;9(4):564-575.
(61) Kim DH, Lipke EA, Kim P, Cheong R, Thompson S, Delannoy M, et al. Nanoscale cues
regulate the structure and function of macroscopic cardiac tissue constructs
Proc.Natl.Acad.Sci.U.S.A. 2010 Jan 12; 107(2):565-570.
(62) Leung B, Sefton MV. A modular approach to cardiac tissue engineering Tissue Eng.Part A.
2010 May 26.
(63) Narmoneva DA, Vukmirovic R, Davis ME, Kamm RD, Lee RT. Endothelial cells promote
cardiac myocyte survival and spatial reorganization - Implications for cardiac regeneration.
Circulation 2004 AUG 24; 110(8):962-968.
(64) Bhana B, Iyer RK, Chen WL, Zhao R, Sider KL, Likhitpanichkul M, et al. Influence of
substrate stiffness on the phenotype of heart cells Biotechnol.Bioeng. 2010 Apr 15;105(6):11481160.
118
(65) Ross RS, Borg TK. Integrins and the myocardium Circ.Res. 2001 Jun 8;88(11):1112-1119.
(66) Meyvantsson I, Beebe DJ. Cell Culture Models in Microfluidic Systems Annual Review of
Analytical Chemistry (2008) 2008;1(1):423 <last-page> 449.
(67) Chung S, Sudo R, Mack PJ, Wan C, Vickerman V, Kamm RD. Cell migration into scaffolds
under co-culture conditions in a microfluidic platform. Lab on a Chip 2009;9(2):269-275.
(68) Sudo R, Chung S, Zervantonakis IK, Vickerman V, Toshimitsu Y, Griffith LG, et al.
Transport-mediated angiogenesis in 3D epithelial coculture. Faseb J. 2009.
(69) Boheler KR, Czyz J, Tweedie D, Yang HT, Anisimov SV, Wobus AM. Differentiation of
pluripotent embryonic stem cells into cardiomyocytes. Circ.Res. 2002 AUG 9;91(3):189-201.
(70) Alsan BH, Schultheiss TM. Regulation of avian cardiogenesis by Fgf8 signaling
Development 2002 Apr; 129(8):1935-1943.
(71) Barron M, Gao M, Lough J. Requirement for BMP and FGF signaling during cardiogenic
induction in non-precardiac mesoderm is specific, transient, and cooperative Dev.Dyn. 2000
Jun;218(2):383-393.
(72) Rathjen J, Rathjen PD. Mouse ES cells: experimental exploitation of pluripotent
differentiation potential Curr.Opin.Genet.Dev. 2001 Oct; 11(5):587-594.
(73) Kim YY, Ku SY, Jang J, Oh SK, Kim HS, Kim SH, et al. Use of long-term cultured
embryoid bodies may enhance cardiomyocyte differentiation by BMP2 Yonsei Med.J. 2008 Oct
31;49(5):819-827.
(74) Takahashi T, Lord B, Schulze PC, Fryer RM, Sarang SS, Gullans SR, et al. Ascorbic acid
enhances differentiation of embryonic stem cells into cardiac myocytes. Circulation 2003 APR
15; 107(14):1912-1916.
(75) Dickman ED, Smith SM. Selective regulation of cardiomyocyte gene expression and cardiac
morphogenesis by retinoic acid Dev.Dyn. 1996 May;206(1):39-48.
(76) Paquin J, Danalache BA, Jankowski M, McCann SM, Gutkowska J. Oxytocin induces
differentiation of P19 embryonic stem cells to cardiomyocytes Proc.Natl.Acad.Sci.U.S.A. 2002
Jul 9;99(14):9550-9555.
(77) Sauer H. Role of reactive oxygen species and phosphatidylinositol 3-kinase in
cardiomyocyte differentiation of embryonic stem cells FEBS Lett. 2000;476(3):218 <last_page>
223.
(78) Hwang YS, Chung BG, Ortmann D, Hattori N, Moeller HC, Khademhosseini A. Microwellmediated control of embryoid body size regulates embryonic stem cell fate via differential
expression of WNT5a and WNT11 Proc.Natl.Acad.Sci.U.S.A. 2009 Oct 6;106(40):16978-16983.
119
(79) Schmelter M, Ateghang B, Helmig S, Wartenberg M, Sauer H. Embryonic stem cells utilize
reactive oxygen species as transducers of mechanical strain-induced cardiovascular
differentiation. FASEB J. 2006 JUN;20(8):1182-+.
(80) Sauer H, Rahimi G, Hescheler J, Wartenberg M. Effects of electrical fields on
cardiomyocyte differentiation of embryonic stem cells J.Cell.Biochem. 1999 Dec 15;75(4):710723.
(81) Shimko VF, Claycomb WC. Effect of Mechanical Loading on Three-Dimensional Cultures
of Embryonic Stem Cell-Derived Cardiomyocytes. Tissue Eng. 0;0(0).
(82) Yu H, Meyvantsson I, Shkel IA, Beebe DJ. Diffusion dependent cell behavior in
microenvironments Lab.Chip 2005 Oct;5(10):1089-1095.
(83) Walker GM, Zeringue HC, Beebe DJ. Microenvironment design considerations for cellular
scale studies Lab.Chip 2004 Apr;4(2):91-97.
(84) Raty S, Walters EM, Davis J, Zeringue H, Beebe DJ, Rodriguez-Zas SL, et al. Embryonic
development in the mouse is enhanced via microchannel culture Lab.Chip 2004 Jun;4(3):186190.
(85) Bratt-Leal AM, Carpenedo RL, McDevitt TC. Engineering the embryoid body
microenvironment to direct embryonic stem cell differentiation Biotechnol.Prog. 2009 JanFeb;25(1):43-51.
(86) Carpenedo RL, Sargent CY, McDevitt TC. Rotary suspension culture enhances the
efficiency, yield, and homogeneity of embryoid body differentiation Stem Cells 2007
Sep;25(9):2224-2234.
(87) Sachlos E, Auguste DT. Embryoid body morphology influences diffusive transport of
inductive biochemicals: a strategy for stem cell differentiation Biomaterials 2008
Dec;29(34):4471-4480.
(88) Bullard TA, Hastings JL, Davis JM, Borg TK, Price RL. Altered PKC expression and
phosphorylation in response to the nature, direction, and magnitude of mechanical stretch.
Can.J.Physiol.Pharmacol. 2007 FEB;85(2):243-250.
(89) Mansour H, de Tombe PP, Samarel AM, Russell B. Restoration of resting sarcomere length
after uniaxial static strain is regulated by protein kinase C epsilon and focal adhesion kinase.
Circ.Res. 2004 MAR 19;94(5):642-649.
(90) Shanker AJ, Yamada K, Green KG, Yamada KA, Saffitz JE. Matrix protein-specific
regulation of Cx43 expression in cardiac myocytes subjected to mechanical load. Circ.Res. 2005
MAR 18;96(5):558-566.
120
(91) Matsuda T, Takahashi K, Nariai T, Ito T, Takatani T, Fujio Y, et al. N-cadherin-mediated
cell adhesion determines the plasticity for cell alignment in response to mechanical stretch in
cultured cardiomyocytes. Biochem.Biophys.Res.Commun. 2005 JAN 7;326(1):228-232.
(92) Lee EJ, Kim DE, Azeloglu EU, Alexandrescu C, Costa KD. Engineered cardiac tissue
chambers demonstrate functional Frank-Starling mechanism and positive stroke work. Circ.Res.
2005 NOV 25;97(11):1200-1200.
(93) Yamane M, Matsuda T, Ito T, Fujio Y, Takahashi K, Azuma J. Raci activity is required for
cardiac myocyte alignment in response to mechanical stress. Biochem.Biophys.Res.Commun.
2007 FEB 23;353(4):1023-1027.
(94) Kada K, Yasui K, Naruse K, Kamiya K, Kodama I, Toyama J. Orientation change of
cardiocytes induced by cyclic stretch stimulation: Time dependency and involvement of protein
kinases. J.Mol.Cell.Cardiol. 1999 JAN;31(1):247-259.
(95) Jacot JG, Martin JC, Hunt DL. Mechanobiology of cardiomyocyte development J.Biomech.
2010 Jan 5;43(l):93-98.
(96) Patwari P, Lee RT. Mechanical control of tissue morphogenesis Circ.Res. 2008 Aug
1;103(3):234-243.
(97) Mammoto T, Ingber DE. Mechanical control of tissue and organ development Development
2010 May;137(9):1407-1420.
(98) Saha S, Ji L, de Pablo JJ, Palecek SP. TGF beta/Activin. Biophys.J. 2008 MAY
15;94(10):4123-4133.
(99) Leor J, Aboulafia-Etzion S, Dar A, Shapiro L, Barbash IM, Battler A, et al. Bioengineered
cardiac grafts: A new approach to repair the infarcted myocardium? Circulation 2000 Nov
7;102(19 Suppl 3):I1156-61.
(100) Kellar RS, Landeen LK, Shepherd BR, Naughton GK, Ratcliffe A, Williams SK. Scaffoldbased three-dimensional human fibroblast culture provides a structural matrix that supports
angiogenesis in infarcted heart tissue Circulation 2001 Oct 23; 104(17):2063-2068.
(101) Ke Q, Yang Y, Rana JS, Chen Y, Morgan JP, Xiao YF. Embryonic stem cells cultured in
biodegradable scaffold repair infarcted myocardium in mice Sheng Li Xue Bao 2005 Dec
25;57(6):673-681.
(102) Samuelson LC. Differentiation of Embryonic Stem (ES) Cells Using the Hanging Drop
Method Cold Spring Harbor Protocols 2006;2006(18):pdb.prot4485 <lastpage> pdb.prot4485.
(103) Matsumoto T, Yung YC, Fischbach C, Kong HJ, Nakaoka R, Mooney DJ. Mechanical
strain regulates endothelial cell patterning in vitro Tissue Eng. 2007 Jan;13(1):207-217.
121
(104) Monzen K, Shiojima I, Hiroi Y, Kudoh S, Oka T, Takimoto E, et al. Bone morphogenetic
proteins induce cardiomyocyte differentiation through the mitogen-activated protein kinase
kinase kinase TAKI and cardiac transcription factors Csx/Nkx-2.5 and GATA-4 Mol.Cell.Biol.
1999 Oct;19(10):7096-7105.
(105) Yamada M, Revelli JP, Eichele G, Barron M, Schwartz RJ. Expression of chick Tbx-2,
Tbx-3, and Tbx-5 genes during early heart development: evidence for BMP2 induction of Tbx2
Dev.Biol. 2000 Dec 1;228(l):95-105.
(106) Heng BC, Haider HK, Sim EKW, Cao T, Ng SC. Strategies for directing the differentiation
of stem cells into the cardiomyogenic lineage in vitro. Cardiovasc.Res. 2004 APR 1;62(l):34-42.
(107) Saha S, Lin J, De Pablo JJ, Palecek SP. Inhibition of human embryonic stem cell
differentiation by mechanical strain. J.Cell.Physiol. 2006 JAN;206(1):126-137.
(108) Chen K, Wu L, Wang ZZ. Extrinsic regulation of cardiomyocyte differentiation of
embryonic stem cells J.Cell.Biochem. 2008 May 1;104(1):119-128.
(109) Gallo P, Condorelli G. Human embryonic stem cell-derived cardiomyocytes: inducing
strategies. Regen.Med. 2006 MAR; 1(2):183-194.
(110) Kehat I, Khimovich L, Caspi 0, Gepstein A, Shofti R, Arbel G, et al. Electromechanical
integration of cardiomyocytes derived from human embryonic stem cells Nat.Biotechnol. 2004
Oct;22(10):1282-1289.
(111) Nussbaum J, Minami E, Laflamme MA, Virag JA, Ware CB, Masino A, et al.
Transplantation of undifferentiated murine embryonic stem cells in the heart: teratoma formation
and immune response FASEB J. 2007 May;21(7):1345-1357.
(112) Baharvand H, Piryaei A, Rohani R, Taei A, Heidari MH, Hosseini A. Ultrastructural
comparison of developing mouse embryonic stem cell- and in vivo-derived cardiomyocytes. Cell
Biol.Int. 2006 OCT;30(10):800-807.
(113) Shapira-Schweitzer K, Habib M, Gepstein L, Seliktar D. A photopolymerizable hydrogel
for 3-D culture of human embryonic stem cell-derived cardiomyocytes and rat neonatal cardiac
cells. J.Mol.Cell.Cardiol. 2009 FEB;46(2):213-224.
(114) Snir M, Kehat I, Gepstein A, Coleman R, Itskovitz-Eldor J, Livne E, et al. Assessment of
the ultrastructural and proliferative properties of human embryonic stem cell-derived
cardiomyocytes Am.J.Physiol.Heart Circ.Physiol. 2003 Dec;285(6):H2355-63.
(115) Sartiani L, Bettiol E, Stillitano F, Mugelli A, Cerbai E, Jaconi ME. Developmental changes
in cardiomyocytes differentiated from human embryonic stem cells: a molecular and
electrophysiological approach. Stem Cells 2007 May;25(5):1136-1144.
122
(116) Liu J, Fu JD, Siu CW, Li RA. Functional sarcoplasmic reticulum for calcium handling of
human embryonic stem cell-derived cardiomyocytes: insights for driven maturation. Stem Cells
2007 Dec;25(12):3038-3044.
(117) Borg TK, Rubin K, Lundgren E, Borg K, Obrink B. Recognition of extracellular matrix
components by neonatal and adult cardiac myocytes Dev.Biol. 1984 Jul;104(1):86-96.
(118) Bullard TA, Borg TK, Price RL. The expression and role of protein kinase C in neonatal
cardiac myocyte attachment, cell volume, and myofibril formation is dependent on the
composition of the extracellular matrix. Microsc.Microanal. 2005 JUN; 11(3):224-234.
(119) Fassler R, Rohwedel J, Maltsev V, Bloch W, Lentini S, Guan K, et al. Differentiation and
integrity of cardiac muscle cells are impaired in the absence of beta 1 integrin J.Cell.Sci. 1996
Dec;109 ( Pt 13)(Pt 13):2989-2999.
(120) Vandenburgh HH, Karlisch P, Farr L. Maintenance of highly contractile tissue-cultured
avian skeletal myotubes in collagen gel In Vitro Cell.Dev.Biol. 1988 Mar;24(3):166-174.
(121) Kofidis T, de Bruin JL, Hoyt G, Ho Y, Tanaka M, Yamane T, et al. Myocardial restoration
with embryonic stem cell bioartificial tissue transplantation J.Heart Lung Transplant. 2005
Jun;24(6):737-744.
(122) Arts T, Costa KD, Covell JW, McCulloch AD. Relating myocardial laminar architecture to
shear strain and muscle fiber orientation. Am.J.Physiol.-Heart Circul.Physiol. 2001
MAY;280(5):H2222-H2229.
(123) Lunkenheimer PP, Redmann K, Kling N, Jiang XJ, Rothaus K, Cryer CW, et al. Threedimensional architecture of the left ventricular myocardium. Anat.Rec.Part A 2006
JUN;288A(6):565-578.
(124) Kaji H, Takii Y, Nishizawa M, Matsue T. Pharmacological characterization of
4
micropatterned cardiac myocytes Biomaterials 2003 Oct;24(23):4239-424 .
(125) Soonpaa MH, Kim KK, Pajak L, Franklin M, Field LJ. Cardiomyocyte DNA synthesis and
binucleation during murine development Am.J.Physiol. 1996 Nov;271(5 Pt 2):H2183-9.
(126) Yamanaka S, Zahanich I, Wersto RP, Boheler KR. Enhanced proliferation of monolayer
cultures of embryonic stem (ES) cell-derived cardiomyocytes following acute loss of
retinoblastoma PLoS One 2008;3(12):e3896.
(127) Sparrow JC, Schock F. The initial steps of myofibril assembly: integrins pave the way
Nat.Rev.Mol.Cell Biol. 2009 Apr; 10(4):293-298.
(128) Sanger JW, Ayoob JC, Chowrashi P, Zurawski D, Sanger JM. Assembly of myofibrils in
cardiac muscle cells Adv.Exp.Med.Biol. 2000;481:89-102; discussion 103-5.
123
(129) Jacot JG, Kita-Matsuo H, Wei KA, Chen HS, Omens JH, Mercola M, et al. Cardiac
myocyte force development during differentiation and maturation Ann.N.Y.Acad.Sci. 2010
Feb; 1188:121-127.
(130) Nishimura S, Yasuda S, Katoh M, Yamada KP, Yamashita H, Saeki Y, et al. Single cell
mechanics of rat cardiomyocytes under isometric, unloaded, and physiologically loaded
conditions Am.J.Physiol.Heart Circ.Physiol. 2004 Jul;287(1):H196-202.
(131) Tanaka Y, Morishima K, Shimizu T, Kikuchi A, Yamato M, Okano T, et al.
Demonstration of a PDMS-based bio-microactuator using cultured cardiomyocytes to drive
polymer micropillars Lab.Chip 2006 Feb;6(2):230-235.
(132) Jonker SS, Zhang L, Louey S, Giraud GD, Thornburg KL, Faber JJ. Myocyte enlargement,
differentiation, and proliferation kinetics in the fetal sheep heart J.Appl.Physiol. 2007
Mar;102(3):1130-1142.
(133) Seki S, Nagashima M, Yamada Y, Tsutsuura M, Kobayashi T, Namiki A, et al. Fetal and
postnatal development of Ca2+ transients and Ca2+ sparks in rat cardiomyocytes.
Cardiovasc.Res. 2003 JUN 1;58(3):535-548.
(134) Fu JD, Yu HM, Wang R, Liang J, Yang HT. Developmental regulation of intracellular
calcium transients during cardiomyocyte differentiation of mouse embryonic stem cells. Acta
Pharmacol.Sin. 2006 JUL;27(7):901-910.
(135) Itzhaki I, Schiller J, Beyar R, Satin J, Gepstein L. Calcium handling in embryonic stem
cell-derived cardiac myocytes: of mice and men. Ann.N.Y.Acad.Sci. 2006 Oct;1080:207-215.
(136) Fisher DJ, Towbin J. Maturation of the heart Clin.Perinatol. 1988 Sep;15(3):421-446.
(137) Clark WA, Decker ML, Behnke-Barclay M, Janes DM, Decker RS. Cell contact as an
independent factor modulating cardiac myocyte hypertrophy and survival in long-term primary
culture J.Mol.Cell.Cardiol. 1998 Jan;30(1):139-155.
(138) Legato MJ. Sarcomerogenesis in human myocardium J.Mol.Cell.Cardiol. 1970
Dec; 1(4):425-437.
(139) Siedner S, Kruger M, Schroeter M, Metzler D, Roell W, Fleischmann BK, et al.
Developmental changes in contractility and sarcomeric proteins from the early embryonic to the
adult stage in the mouse heart J.Physiol. 2003 Apr 15;548(Pt 2):493-505.
(140) Zak R. Development and proliferative capacity of cardiac muscle cells Circ.Res. 1974
Aug;35(2):suppl II:17-26.
(141) Caspi 0, Itzhaki I, Kehat I, Gepstein A, Arbel G, Huber I, et al. In vitro
electrophysiological drug testing using human embryonic stem cell derived cardiomyocytes Stem
Cells Dev. 2009 Jan-Feb; 18(1):161-172.
124
(142) Reppel M, Pillekamp F, Brockmeier K, Matzkies M, Bekcioglu A, Lipke T, et al. The
electrocardiogram of human embryonic stem cell-derived cardiomyocytes J.Electrocardiol. 2005
Oct;38(4 Suppl): 166-170.
(143) Xi J, Schmidt JJ, Montemagno CD. Self-assembled microdevices driven by muscle
Nat.Mater. 2005 Feb;4(2):180-184.
(144) Radisic M, Euloth M, Yang L, Langer R, Freed LE, Vunjak-Novakovic G. High-density
seeding of myocyte cells for cardiac tissue engineering Biotechnol.Bioeng. 2003 May
20;82(4):403-414.
(145) Souren JE, Schneijdenberg C, Verkleij AJ, Van Wijk R. Factors controlling the rhythmic
contraction of collagen gels by neonatal heart cells In Vitro Cell.Dev.Biol. 1992 Mar;28A(3 Pt
1):199-204.
(146) Yamamura N, Sudo R, Ikeda M, Tanishita K. Effects of the mechanical properties of
collagen gel on the in vitro formation of microvessel networks by endothelial cells Tissue Eng.
2007 Jul;13(7):1443-1453.
(147) Bhana B, Iyer RK, Chen WL, Zhao R, Sider KL, Likhitpanichkul M, et al. Influence of
substrate stiffness on the phenotype of heart cells Biotechnol.Bioeng. 2010 Apr 15;105(6):11481160.
(148) Jacot JG, McCulloch AD, Omens JH. Substrate stiffness affects the functional maturation
of neonatal rat ventricular myocytes Biophys.J. 2008 Oct;95(7):3479-3487.
(149) Engler AJ, Griffin MA, Sen S, Bonnemann CG, Sweeney HL, Discher DE. Myotubes
differentiate optimally on substrates with tissue-like stiffness: pathological implications for soft
or stiff microenvironments J.Cell Biol. 2004 Sep 13;166(6):877-887.
(150) Discher DE, Janmey P, Wang YL. Tissue cells feel and respond to the stiffness of their
substrate. Science 2005 NOV 18;310(5751):1139-1143.
(151) Simpson DG, Terracio L, Terracio M, Price RL, Turner DC, Borg TK. Modulation of
cardiac myocyte phenotype in vitro by the composition and orientation of the extracellular
matrix J.Cell.Physiol. 1994 Oct; 161(1):89-105.
(152) Yu JG, Russell B. Cardiomyocyte remodeling and sarcomere addition after uniaxial static
strain in vitro J.Histochem.Cytochem. 2005 Jul;53(7):839-844.
(153) Guan K, Czyz J, Furst DO, Wobus AM. Expression and cellular distribution of
alpha(v)integrins in beta(1)integrin-deficient embryonic stem cell-derived cardiac cells
J.Mol.Cell.Cardiol. 2001 Mar;33(3):521-532.
125
(154) Chen TT, Luque A, Lee S, Anderson SM, Segura T, Iruela-Arispe ML. Anchorage of
VEGF to the extracellular matrix conveys differential signaling responses to endothelial cells
J.Cell Biol. 2010 Feb 22;188(4):595-609.
(155) Bird SD, Doevendans PA, van Rooijen MA, de la Riviere AB, Hassink RJ, Passier R, et al.
The human adult cardiomyocyte phenotype. Cardiovasc.Res. 2003 MAY 1;58(2):423-434.
(156) Radisic M, Park H, Shing H, Consi T, Schoen FJ, Langer R, et al. Functional assembly of
engineered myocardium by electrical stimulation of cardiac myocytes cultured on scaffolds
Proc.Natl.Acad.Sci.U.S.A. 2004 Dec 28;101(52):18129-18134.
(157) Fink C, Ergun S, Kralisch D, Remmers U, Weil J, Eschenhagen T. Chronic stretch of
engineered heart tissue induces hypertrophy and functional improvement FASEB J. 2000
Apr; 14(5):669-679.
(158) Park J, Kim IC, Baek J, Cha M, Kim J, Park S, et al. Micro pumping with cardiomyocytepolymer hybrid Lab.Chip 2007 Oct;7(10):1367-1370.
(159) Rogers SS, Waigh TA, Zhao X, Lu JR. Precise particle tracking against a complicated
background: polynomial fitting with Gaussian weight Phys.Biol. 2007 Oct 9;4(3):220-227.
(160) Yamamoto K, Dang QN, Maeda Y, Huang H, Kelly RA, Lee RT. Regulation of
cardiomyocyte mechanotransduction by the cardiac cycle Circulation 2001 Mar
13;103(10):1459-1464.
(161) Sugiura S, Nishimura S, Yasuda S, Hosoya Y, Katoh K. Carbon fiber technique for the
investigation of single-cell mechanics in intact cardiac myocytes Nat.Protoc. 2006;1(3):14531457.
(162) Lin G, Palmer RE, Pister KS, Roos KP. Miniature heart cell force transducer system
implemented in MEMS technology IEEE Trans.Biomed.Eng. 2001 Sep;48(9):996-1006.
(163) Berk BC, Fujiwara K, Lehoux S. ECM remodeling in hypertensive heart disease
J.Clin.Invest. 2007 Mar; 117(3):568-575.
(164) Gallagher GL, Jackson CJ, Hunyor SN. Myocardial extracellular matrix remodeling in
ischemic heart failure Front.Biosci. 2007 Jan 1;12:1410-1419.
(165) WRITING GROUP MEMBERS, Lloyd-Jones D, Adams RJ, Brown TM, Carnethon M,
Dai S, et al. Heart disease and stroke statistics--2010 update: a report from the American Heart
Association Circulation 2010 Feb 23;121(7):e46-e215.
(166) Dispersyn GD, Geuens E, Ver Donck L, Ramaekers FC, Borgers M. Adult rabbit
cardiomyocytes undergo hibernation-like dedifferentiation when co-cultured with cardiac
fibroblasts Cardiovasc.Res. 2001 Aug 1;51(2):230-240.
126
(167) Passier R, van Laake LW, Mummery CL. Stem-cell-based therapy and lessons from the
heart Nature 2008 May 15;453(7193):322-329.
(168) Zhang F, Pasumarthi KB. Embryonic stem cell transplantation: promise and progress in the
treatment of heart disease BioDrugs 2008;22(6):361-374.
(169) METZGER JM, LIN WI, SAMUELSON LC. Transition in Cardiac Contractile Sensitivity
to Calcium during the In-Vitro Differentiation of Mouse Embryonic Stem-Cells. J.Cell Biol.
1994 AUG;126(3):701-711.
(170) Lee P, Lin R, Moon J, Lee LP. Microfluidic alignment of collagen fibers for in vitro cell
culture. Biomed.Microdevices 2006;8(1):35-41.
(171) Verdu E, Labrador RO, Rodriguez FJ, Ceballos D, Fores J, Navarro X. Alignment of
collagen and laminin-containing gels improve nerve regeneration within silicone tubes.
Restorative Neurol.Neurosci. 2002;20(5):169-179.
(172) Kim C, Majdi M, Xia P, Wei KA, Talantova M, Spiering S, et al. Non-cardiomyocytes
influence the electrophysiological maturation of human embryonic stem cell-derived
cardiomyocytes during differentiation Stem Cells Dev. 2010 Jun; 19(6):783-795.
(173) Livoti CM, Morgan JR. Self-assembly and tissue fusion of toroid-shaped minimal building
units Tissue Eng.Part A. 2010 Jun;16(6):2051-2061.
(174) Zamanian B, Masaeli M, Nichol JW, Khabiry M, Hancock MJ, Bae H, et al. Interfacedirected self-assembly of cell-laden microgels Small 2010 Apr 23;6(8):937-944.
127