University of Southern California Dissertations and Theses

Engineering scalable two- and three-dimensional striated muscle microtissues for human disease modeling

(USC Thesis Other)

Engineering scalable two- and three-dimensional striated muscle microtissues for human disease modeling

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content 1

ENGINEERING SCALABLE TWO- AND THREE-
DIMENSIONAL STRIATED MUSCLE MICROTISSUES
FOR HUMAN DISEASE MODELING

BY

NETHIKA R. ARIYASINGHE

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
of
THE UNIVERSITY OF SOUTHERN CALIFORNIA

In partial fulfillment of the requirements for the degree of
DOCTOR OF PHILOSOPHY
In
BIOMEDICAL ENGINEERING

AUGUST 2019
2

Table of Contents
Abstract ........................................................................................................................................... 7
1 Introduction: Engineered Tissues as Next-Generation Models for Striated Muscle Disease
Modeling and Drug Testing ............................................................................................................ 8
1.1 Striated Muscle Physiology and Pathophysiology ........................................................... 8
1.1.1 Structure and Function of Healthy Myocardial Tissue ................................................. 9
1.1.2 Structure and Function of Diseased Myocardial Tissue ............................................. 10
1.1.3 Structure and Function of Healthy Skeletal Muscle Tissue ....................................... 10
1.1.4 Structure and Function of Diseased Skeletal Muscle Tissue ...................................... 11
1.2 Bottlenecks in the Drug Development Process .............................................................. 12
1.2.1 Cell Culture Models.................................................................................................... 13
1.2.2 Cardiac Muscle Cell Culture Models ......................................................................... 13
1.2.3 Skeletal Muscle Cell Culture Models ......................................................................... 15
1.2.4 Striated Muscle Animal Models ................................................................................. 16
1.2.5 Animal Models for Cardiac Disease Modeling and Drug Testing ............................. 16
1.2.6 Animal Models for Skeletal Muscle Disease Modeling and Drug Testing ................ 18
1.3 Engineered Tissues as Next-Generation Disease Models .............................................. 19
1.3.1 Existing 2D Engineered Striated Tissue Models ........................................................ 20
1.3.2 Existing 3D Engineered Tissue Models ..................................................................... 21
1.4 Objective and Aims ........................................................................................................ 23
2 Aim 1: Engineer a scalable platform to test the effects of ECM and cellular remodeling on
the contractility of 2D engineered cardiac tissues. ....................................................................... 25
3 Aim 2: Engineer more scalable 3D human skeletal muscle tissues using 3D-printing for rapid
prototyping and tissue clearing for enhanced visualization .......................................................... 50
4 Conclusion .............................................................................................................................. 72
4.1 Scalable 2D engineered cardiac tissues demonstrate the effects of disease-relevant
microenvironmental factors on contractile phenotype .............................................................. 73
4.2 The throughput and visualization of 3D engineered skeletal muscle tissues is improved
using 3D-printing and tissue clearing ........................................................................................ 75
4.3 Advantages and disadvantages of 2D and 3D systems .................................................. 78
4.4 Organs-on-Chips in the Future ....................................................................................... 79
References ..................................................................................................................................... 81
Supplemental 1: Aim 2 ................................................................................................................. 88
Supplemental 2: Aim 3 ................................................................................................................. 99
3

Figures

Figure 1. Sarcomeres are the Basic Functional Unit of Striated Muscle ................................. 8

Figure 2. The Structure of Myocardial Tissue. .......................................................................... 9

Figure 3. The Structure of Skeletal muscle .............................................................................. 11

Figure 4. A Comparison of Normal and DMD Skeletal Muscle ............................................. 11

Figure 5. Timeline of Drug Discovery and Development. ....................................................... 12

Figure 6: Fabrication of µMyocardium .................................................................................... 34

Figure 7: Cell demographics and actin alignment in engineered µMyocardia. .................... 35

Figure 8: Quantifying contractile stresses generated by µMyocardia ................................... 37

Figure 9: Displacement and traction stress vector colormaps for multiple µMyocardia .... 38

Figure 10: Spatial distributions of x-displacement vectors in µMyocardia .......................... 39

Figure 11: Average longitudinal contractile parameters for µMyocardia ............................ 40

Figure 12: Expression of non-patterned ECM in µMyocardia............................................... 42

Figure 13: Fabrication of Muscle Bundle Templates .............................................................. 59

Figure 14: Engineered muscle bundle survival and compaction. ........................................... 60

Figure 15: Contractile response of engineered muscle bundles to electrical stimulation. ... 62

Figure 16: Images of cyrosectioned and immunostained engineered muscle bundles ......... 63

Figure 17: Schematic of tissue clearing ..................................................................................... 64

Figure 18: Comparison of uncleared and cleared muscle bundles......................................... 64

Figure 19: Dystrophin and α-actinin in cleared engineered muscle bundles ........................ 65

Figure 20: Cross-sectional confocal slices of uncleared and cleared immunostained muscle
bundles ......................................................................................................................................... 66

Figure 21: Overview of the fabrication and analysis of 2D engineered MicroMyocardium.
....................................................................................................................................................... 73
4

Figure 22: Overview of fabrication and analysis of 3D engineered skeletal muscle bundles
....................................................................................................................................................... 76

Figure 23: The place of organ-on-a-chip as model systems. ................................................... 80

5

Acknowledgements
It has truly taken a village to get to this point. First of all, I would like to thank Dr.
Megan McCain for accepting me into your lab. Thank you for facilitating my journey as Ph.D.
student and for being not only an advisor but also a mentor. Your ability to tailor your mentoring
style to each student is truly a special gift. It’s been a pleasure to watch you grow into your role
as a professor and to be a part of the great lab you’ve created. I would also like to thank my
committee members, Drs. David D’Argenio, Jill McNitt-Gray, Eun Ji Chung, and Keyue Shen
for their time and feedback. Thank you also to Dr. Stacey Finley for your encouragement on this
journey. Thank you to Mark Pincus, Terri Suzuki, and Paul August for your support and
guidance during my time at the Sanofi Tucson Research Center. Zach Remer, I appreciated your
company and enjoyed working with you there. Thank you also to Mischal Diasanta and William
Yang, the former and current BME department Graduate Affairs Advisors. You both have done
such a great job directing us Ph.D. students with great attitudes and humor.
Thank you to my former and current labmates. Jasper Hsu, I learned so much from you.
Davi Lyra-Leite, it’s been a pleasure going through this whole journey with you, and we have
both grown so much! Nathan Cho, thank you for the food recommendations and your “Nacho
Quotes” that brightened my day. Thank you also to Andrew Peterson and Jeffrey Santoso for
your scientific input and friendly conversation. Patrick Vigneault and Megan Rexius-Hall, it’s
been a pleasure working with y’all, and thank you for your advice on research and life. Thank
you also to the students I worked with, including Archana Bettadapur, Gio Suh, Alyssa Viscio,
and Caitlin Reck.
I’ve been lucky to meet some really great people in this department, including Jen Rohrs,
Adam Mergenthal, Alex Baldwin, and Marilena Dimotsantou, in few. Thanks also to my friends
6

Michele Nieves, Kelly Knack, Gabrielle Yee, and Karen Fang for your support and
entertainment. Thank you Aubrey Graham for the beats that helped me pass time while doing
labwork, and thank you Jordan KW for providing indispensable distraction when I needed it as
well as reminding me that every day is different and not to be too hard on myself.
Last but certainly not least, thank you so so much to Uldric Antao and my family. Uldric,
I know my weekend cell culture media changes also became your weekend media changes, and I
appreciate your support and encouragement. To my parents, thanks for shaping me into the
person I am today, fueling my scientific curiosity and creativity, and putting up with my desire to
climb wherever looks fun (figuratively and literally). Going through a PhD program has made
me even more aware of the sacrifices you both have made to bring us to where we are today.
Thanks to my sister Nethmi. We don’t always see the world the same way, but you have played
an incredible role in shaping my life from the first moment you helped me put the rings on my
Fisher Price toy appropriately.
7

Abstract
Striated muscle diseases, including cardiovascular diseases and muscular dystrophies, are a
significant burden on the United States. Our ability to develop therapies for these diseases is
hampered by a reliance on models systems that cannot accurately and efficiently replicate human
striated muscle tissue and/or quantify its phenotype. Also, striated muscle diseases can be either
acquired or genetic, which present unique challenges when modeling these different forms. To
overcome these challenges, tissue engineering has recently been utilized to develop more
sophisticated models of striated muscle, but many of these models still lack physiological
relevance, scalability, reproducibility, and functional outputs. Here, our goal was to engineer
scalable striated muscle tissue platforms that can be used to determine the effects of disease-
relevant microenvironmental factors and/or genetic mutations on muscle phenotype. First, we
engineered a 2D platform to quantify the effects of cellular and extracellular remodeling on the
contractility of 2D cardiac tissues. We will also engineer 3D skeletal muscle tissues using 3D-
printing for rapid prototyping and tissue clearing for enhanced visualization. We established our
approach using C2C12 mouse myoblasts by determining the effect of mold dimensions on the
percentage survival and change in bundle width of 3D bundles. We also demonstrated that
bundles respond to electrical stimulation by contracting. Last, we used tissue clearing to
visualize the bundles in multiple planes. The platforms developed here can be used to create
“organs-on-chips” mimicking the structure and function of the heart and skeletal muscle. These
chips could be used to create high-throughput quantification of organ function and organ
interactions to drugs.

8

1 Introduction: Engineered Tissues as Next-Generation Models for Striated
Muscle Disease Modeling and Drug Testing
Diseases that afflict striated muscle (heart and skeletal muscle) annually affect many
people in the United States. For example, cardiovascular diseases are the leading cause of death
in the United States
2
. While there are a variety of causes for cardiovascular diseases, most of
them, including myocardial infarctions, are acquired. Conversely, skeletal muscle disorders are
primarily inherited. For example, Duchenne Muscular Dystrophy (DMD) affects 1 in every
3,500 male births
3
and causes debilitating muscle degeneration and severe loss of mobility at a
relatively young age. One reason why these diseases remain incurable is that existing model
systems for disease modeling and drug screening do not accurately replicate the complexities of
healthy and diseased human striated muscle tissues. These limitations have significantly hindered
the drug development process and contribute to the ongoing burden of striated muscle diseases.
In this chapter, we will review the structure and function of cardiac and skeletal muscle, and
summarize existing approaches to model these diseases. We will then describe our approach to
engineer two-dimensional and three-
dimensional scalable human striated muscle
tissue platforms that can advance our ability to
determine the effects of disease-relevant
microenvironmental factors and genetic
mutations on muscle phenotype.
1.1 Striated Muscle Physiology and
Pathophysiology
Cardiac and skeletal muscle tissue have
many similarities in their basic structure and
function. For example, both tissues consist of striated muscle cells, known as myocytes, that
Figure 1. Sarcomeres are the Basic Functional Unit of
Striated Muscle. An example of striated muscle, skeletal
muscle, is shown here. Courtesy of Encyclopaedia
Britannica.
9

shorten and generate force. These cells contain sarcomeres with repeating units of actin and
myosin filaments between two Z-lines (Error! Not a valid bookmark self-reference.). The
globular head of the thick myosin filaments binds to the thin actin filaments, which are attached
to the Z-lines by α -actinin, causing sarcomeric shortening. This sarcomeric shortening results in
muscle contraction and force generation. However, they also have many key differences that
must be considered when developing healthy and diseased engineered tissue models. Here, we
outline the key structural and functional characteristics of healthy and diseased cardiac and
skeletal muscle tissue.
1.1.1 Structure and Function of Healthy Myocardial Tissue
The heart pumps blood due to the repeated contractions of the muscular tissue of the
ventricles, known as myocardium. Myocardium consists primarily of aligned cardiac myocytes
and fibroblasts embedded in an extracellular
matrix (ECM) network. In native tissue, cardiac
myocytes are aligned in parallel (Figure 2) and
contract in unison, leading to force generation
parallel to the alignment of the cells
4
. Fibroblasts
are interspersed among the aligned cardiac
myocytes and play a mostly supportive role by
synthesizing, depositing, and degrading the ECM
5
.
Myocardial ECM contains collagen, fibronectin,
and laminin. Collagen forms fibers that
structurally stabilize the myocardium, and cardiac myocytes and fibroblasts are anchored to these
collagen fibers by the basement membrane, composed mostly of the glycoproteins fibronectin
(FN) and laminin (LN)
4, 6
. Thus, healthy myocardial tissue consists of aligned cardiac myocytes
Figure 2. The Structure of Myocardial Tissue.
Cardiac tissue contains sarcomeric cardiac myocytes
that are highly aligned and embedded in a
biochemically- and mechanically-defined ECM.
Image courtesy of Stephen Gallik at
http://histologyolm.stevegallik.org/node/146.

10

and fibroblasts surrounded by a mechanically- and biochemically-defined ECM that is critical for
its structure and function. Mimicking this cellular and extracellular architecture is crucial for
engineering physiologically-relevant models of myocardial tissue for disease modeling and drug
testing.
1.1.2 Structure and Function of Diseased Myocardial Tissue
Many cardiac diseases are associated with remodeling of both the cellular and
extracellular components of the myocardium. For example, a myocardial infarction occurs when
an arterial blockage causes myocyte necrosis. To allow immune cells to remove the necrotic
myocytes, proteases degrade the existing ECM. Next, fibroblasts migrate to the area and deposit
new ECM as scar tissue
7
. The ECM of post-infarct myocardium is stiffer due to an increase in
collagen and also contains less fibronectin and laminin than healthy myocardium
8
. ECM
remodeling can also be caused by conditions such as tachyarrhythmia, a rapid heart rate.
Tachyarrhythmia results in the loss of normal ECM collagen content and distribution and
diminished myocyte attachment to the basement membrane, which causes a loss of normal ECM
support and architecture as well as alterations in myocyte support and alignment
9-12
. This
remodeling can lead to modified function of the left ventricle and eventually to heart failure.
Thus, remodeling of the ECM occurs during many cardiac diseases, and understanding the
effects of this remodeling on the myocardium phenotype could lead to more effective treatments
for cardiac disease.
1.1.3 Structure and Function of Healthy Skeletal Muscle Tissue
Skeletal muscle is composed of bundles of aligned muscle fibers embedded in a network
of collagen I fibers (Figure 3) that are surrounded by other proteoglycans and glycoproteins,
including other types of collagen, which maintain the structure and organization of the ECM
13
.
Muscle fibers are formed by the differentiation of precursor cells known as skeletal myoblasts,
11

which fuse and mature to form multi-nucleated, striated muscle fibers. Skeletal muscle contracts
when sarcomeres in muscle fibers shorten in unison,
parallel to the fibers. This allows humans to engage in
voluntary movements. Due to this close coupling between
structure and function, mimicking the alignment of human
skeletal muscle fibers and the composition of the ECM is
crucial for engineering relevant models of skeletal muscle
tissue.
1.1.4 Structure and Function of Diseased Skeletal
Muscle Tissue
Some of the most common diseases that affect
skeletal muscle are the muscular dystrophies. Muscular
dystrophies are inherited myopathies caused by a variety
of genetic mutations that lead to skeletal muscle
degeneration and loss of function
14
. The most common
muscular dystrophy is Duchenne Muscular Dystrophy (DMD). In this disease, a mutation in the
DMD gene interferes with the production of dystrophin, a protein that connects muscle fibers to
the ECM (Figure 4). The lack of dystrophin compromises the structure of the muscle fiber and
weakens the membrane, leading to
progressive muscle degeneration and loss of
proper function
15
.
Another class of muscular dystrophy
is Limb-girdle Muscular Dystrophies. These
are classified into 15 variations, each of
Figure 3. The Structure of Skeletal
muscle. Skeletal muscle contains bundles
of sarcomeric musle fibers surrounded by
ECM. Image courtesy of Gillies, AR
Muscle Nerve 2011.
Figure 4. A Comparison of Normal and DMD Skeletal
Muscle. The skeletal muscle of a DMD patient lacks the
dystrophin seen in normal human skeletal muscle. Image
Courtesy of Blake et al, Physiological Reviews, 2002.
12

which is caused by a different genetic mutation that affects a certain protein, leading to
compromised muscle structure and weakened muscle function. For example, limb-girdle
muscular dystrophy types 2C, 2D, 2E, and 2F involve mutations in one of the sarcoglycan genes,
which destabilizes the sarcoglycan complex at the plasma membrane, causing an inability to
counteract the mechanical stress of muscle contractions
16, 17
. Limb-girdle muscular dystrophy
type 2A occurs due to mutations in the calpain3 gene, disrupting muscle membrane homeostasis
18, 19
. Thus, because there are many diverse forms of muscular dystrophy each associated with
different genetic mutations, it has been challenging to rapidly generate models of dystrophic
skeletal muscle that sufficiently recapitulate the unique genotypes and phenotypes of these
debilitating diseases.
1.2 Bottlenecks in the Drug Development Process
Drug discovery and development today is relatively low-throughput, inefficient, and
expensive (Figure 5). During the initial drug discovery phase, researchers screen five to ten
thousand compounds, from which approximately five will enter clinical trials, and only one will
eventually become FDA-approved. It can take three to six years for a drug to pass from the drug
discovery phase into clinical
trials, and then another six to
seven years to pass from
clinical trials to FDA
review. After another one to
two years of review and
manufacturing, the drug will
be available on the market.
Figure 5. Timeline of Drug Discovery and Development.
Courtesy of focr.org.
13

Despite the 10-year process used to develop drugs, unforeseen side effects and toxicity
have still been observed
20-22
, leading to market withdrawal of FDA-approved drugs
23
. One reason
for these issues is that existing model systems, including cell culture and animal models, fall
short in recapitulating native human tissue responses. To overcome these shortcomings,
engineered human tissue models have emerged as a new approach to drug screening. Here, we
summarize existing approaches to modeling healthy and diseased striated muscle tissue,
including newly-developed engineered tissues.
1.2.1 Cell Culture Models
Conventional cell culture entails isolating and maintaining striated muscle cells on glass
or polystyrene surfaces, usually coated with ECM proteins. Cell culture is a relatively
inexpensive and high-throughput approach to drug testing. However, these platforms lack the
native architecture and complexity of native tissues, such as alignment, and cannot be used to
determine all pathophysiological mechanisms, such as ECM remodeling. Thus, although cell
culture models are high-throughput, they do not replicate native tissue structure or
microenvironment and cannot be used to quantify the function of striated muscle tissues or
organs, limiting human relevance of these models.
1.2.2 Cardiac Muscle Cell Culture Models
Because adult cardiac myocytes are post-mitotic, it is not possible to develop and
propagate a cell line of cardiac myocytes. Isolating primary human cardiac myocytes is also not
feasible. Thus, primary cardiac myocytes isolated from animals (typically rodents, such as
neonatal rats) and human stem cell-derived cardiac myocytes are the only practical options for
cardiac cell culture models. Primary neonatal rat cardiac myocytes are routinely used because
they are an inexpensive, easily-acquired cell type. However, these rodent cells cannot directly
replicate human function and toxicity. For example, a rodents’ heart beat is faster than a human
14

heart beat. Thus, drugs for human arrhythmias often cannot be accurately tested using rat cardiac
myocytes
24
. There are also multiple types of stems cells that have been used in cell culture
models: embryonic stem cells, mesenchymal stem cells, and induced pluripotent stem cells
(iPSCs). Embryonic stem cells are isolated from the blastocyst during embryonic development
25
.
Mesenchymal stem cells primarily reside in the bone marrow, but they may also be located in all
postnatal organs
26-28
. Both embryonic and mesenchymal stem cells are not easily accessible,
limiting usage and throughput. Using embryonic stem cells also raises ethical issues. iPSCs are a
relatively new type of stem cell generated by reprogramming adult cells, such as fibroblasts, into
a pluripotent state. These cells can then be differentiated into another type of cell, including
cardiac myocytes. Human stem cell-derived cardiac myocytes have potential to mimic human
responses more closely than animal cells. Importantly, iPS cells could also be used for
personalized medicine, as a sample of skin fibroblasts taken from a patient could be used to
generate iPSC-derived cardiac myocytes. Previous work has shown that the effect of cardioactive
drugs on iPSC-derived cardiac myocytes is similar to empirical results from clinical trials
24
.
However, despite the benefits of stem cell-derived cardiac myocytes, conventional cell culture
approaches still lack the highly-aligned architecture and biochemically- and mechanically-
defined microenvironment of native cardiac tissue, limiting the overall relevance of these
platforms.
Most assays used with cardiac myocytes in culture measure indirect proxies of
contractility, such as cell toxicity, myocyte beating rate, and electrical potential
24, 29, 30
.
Contractility has been quantified in previous work based on the change in myocyte length from
diastole to systole
24
. While change in myocyte length can quantify contractility, this method is
unsophisticated and relies on the researcher manually determining the lengths at diastole and
15

systole. Last, the electrical potential of cardiac myocytes cannot determine strength of tissue
communication as this does not quantify electrical signal propagation through the cells. Thus,
conventional cardiac cell culture models not only lack the architecture and the microenvironment
of native cardiac tissue, but also cannot quantify the contractile phenotype of aligned native
cardiac tissues.
1.2.3 Skeletal Muscle Cell Culture Models
Unlike cardiac myocytes, skeletal myoblasts do proliferate, allowing primary rat skeletal
myoblasts, skeletal myoblast cell lines, and human skeletal myoblasts to be cultured using
traditional cell culture methods
31,32-35
. Cell lines are attractive because they are a cheap, easily-
obtainable option. The mouse-derived myoblast C2C12 line is particularly attractive because it
differentiates rapidly. While the majority of cell lines are derived from mice or rats, there is the
human RCMH cell line
34
which is human-relevant and contains muscle-specific ionic channels
and IP3 receptors that could be used as drug targets to control skeletal muscle contraction or
relaxation
32
. For many lines and primary myoblasts, the extent of cell fusion can decline with
passage
32
, and myogenin is not consistently expressed, as some subsets of the gene are only seen
after prolonged cultivation
36-38
. Also, traditional cell culture substrates cannot be used for long-
term culture of skeletal muscle because, after approximately a week, the myotubes will
delaminate from matrix-coated surfaces
39-42
. This prevents skeletal muscle cell cultures from
being used for long-term disease modeling. Because most of the skeletal muscle cells cultured
are not human, lack the gene expression of native tissues, and lack the architecture and
microenvironment of native skeletal muscle, their ability to replicate healthy and pathological
human skeletal muscle is limited.
Previous approaches to characterizing skeletal muscle in culture have included measuring
gene expression, imaging calcium transients, and quantifying force with traction force
16

microscopy. Analysis of gene expression has been used to establish the ion channel
concentration, maturity, and renewability of skeletal muscle
32, 36, 37,35, 38, 43
, while imaging
calcium transients quantifies electrophysiological function
44
. Gene expression is by far the most
popular method of quantifying the phenotype of skeletal muscle, but this is an indirect indicator
of skeletal muscle structure and function. Skeletal muscle function can be more directly
quantified by using traction force microscopy to determine force generation, but this cannot be
used in long-term cultures due to delamination
33
. Therefore, existing skeletal muscle cell culture
models cannot quantify long-term skeletal muscle phenotype due to the inability to culture the
cells, and current assays also only give indirect indicators of skeletal muscle function.
1.2.4 Striated Muscle Animal Models
Animal models are advantageous for investigating disease mechanisms and the effects of
drugs and cell- and gene-based therapies on animal survival and organ structure and function in
an intact organism. Drawbacks to these models include high cost, limited relevance to humans,
and low throughput
45-47
. Furthermore, the utility of drug targets or disease mechanisms identified
in animals is limited because the researcher is extrapolating that the response or mechanism seen
in an animal is similar to the human response or mechanism. Due to genetic and anatomical
differences, as well as differences in physiological and pathophysiological mechanisms, findings
in animal results may differ from the human response or mechanisms. Even though animal
models can be used to examine disease mechanisms and drug effects on cardiac and skeletal
muscle, these results have limited relevance to humans.
1.2.5 Animal Models for Cardiac Disease Modeling and Drug Testing
Zebrafish, small mammals, and large mammals have previously been used for cardiac
disease modeling and drug testing. Zebrafish are an attractive model for high-throughput
screening because zebrafish embryos are small enough to be raised in 96- or 384-well plates and
17

can absorb drugs that are placed in the media around the embryo
48, 49
. Because zebrafish are
transparent, digital motion analysis can be used to examine cardiac output, vessel diameter, and
heart rate
48, 50
. Small mammal models include mice, rats, guinea pigs, and rabbits. A model for
cardiac disease can be created fairly easily in these animals by occluding a coronary artery
51
, and
then this model can be used to examine pathophysiological progression, or to test drugs
administered by injection
46
. Examining ex vivo tissue structure through imaging techniques, such
as electron microscopy, allows researchers to determine changes to cardiac tissue structure, such
as loss of alignment, or toxicity
46
. Large mammal models, such as dogs, can also be surgically-
modified to create models of cardiac disease and given drugs by injection. Rate of animal
survival and histological examination similar to the methods used for small mammals are used to
determine drug success
47
.
The utility of animal models is limited by variation between species and our inability to
control many aspects of the tissue microenvironment in animals. Zebrafish proteins are less than
70% similar to their human orthologues
52
. Variation in cardiac structure may result in different
physiological mechanisms in animals
53
or may also preclude the reproduction of certain human
cardiac diseases
54
. Even if certain diseases are reproduced, diseases such as arrhythmias may be
easier to reverse in smaller mammals due to the smaller size of their hearts
55
. Furthermore, many
human cardiac diseases, such as infarction, result in environmental changes to the cardiac tissue.
While animal models can be used to investigate the progression from healthy to diseased tissue,
they cannot be used to determine how individual changes to the tissue microenvironment, such
as changing the elasticity or protein composition of the ECM, impacts disease progression. This
limits our ability to identify possible drug targets. Variability between species and inability to
18

control cardiac tissue microenvironment limit the utility of animal models in cardiac disease
modeling and drug testing.
1.2.6 Animal Models for Skeletal Muscle Disease Modeling and Drug Testing
Common animal models used to model skeletal muscle and muscular dystrophy include
zebrafish, mice, rats, and dogs. Due to similarities between zebrafish and mammalian skeletal
muscle, zebrafish have been used to investigate skeletal muscle repair and regeneration
56-59
.
Zebrafish structure can also be studied using various dyes, stains, or transgenic skeletal muscle
fluorescent reporters as well as the birefringence assay, which allows the user to examine
changes in sarcomere structure
60-63,64
. Zebrafish skeletal muscle health can be quantified based
on animal survival
63,64
, as well as more functional measurements include spontaneous embryo
coiling and larval touch-evoked escape response
61, 65-67
, electrophysiological recordings
65, 68
, or
direct measurement of contractile properties using a force transducer
69
. The transgenic mdx mice
are a popular model for DMD
70, 71
. Muscle structure can be determined using histological
examination of muscle for muscle fiber necrosis, cellular infiltration of inflammatory cells,
regeneration, and fibrosis
72, 73
. Functional assays for mdx mice include examination of twitch
contraction kinetics in limb and diaphragm muscles, capacity for exercise, spontaneous
locomotor activity, mouse strength, and electrophysiological recordings
74-77
. However, because
mdx mice only show moderate signs of degeneration, the relevance of results obtained with mdx
mice is limited. The golden retriever muscular dystrophy (GRMD) model, in contrast, develops a
progressive, fatal condition similar to human muscular dystrophy patients
78, 79
. Joint
contractures
80
, muscle weakness
81, 82
, and timed function tests are used to record disease
progression in GRMD models. Measurement of tibiotarsal force and torque can also assess
contractile function in both mdx mice and GRMD models
83
. Thus, animal models have many
advantages for understanding disease progression in an intact organism.
19

The limitations of skeletal muscle animal models are similar to those for cardiac animal
models: there are genetic, physiological and pathophysiological differences between humans and
animal models, and the human-relevant, large mammal models are fairly low-throughput and
expensive. Zebrafish can be polyploid for specific genes, making it difficult to identify the
effects of genetic mutations and to generate specific mutations
84, 85
. Zebrafish regenerate muscle
from stem-cell-like cells, not from de-differentiated fibers, which may lead to differences in
muscle regeneration physiology
86
. As mentioned above, the mdx mouse undergoes less muscle
degeneration and regeneration than dystrophin-null humans and contains less fibrosis and fewer
inflammatory cells
87-90
, meaning that drug and cell-based therapies that are successful in the mdx
mouse may not be successful in the more debilitating human muscular dystrophies. While the
GRMD model is a better model for MD in humans than the mdx mouse, dogs are a low-
throughput and costly model. Dogs have smaller litters and a long generation time and require
significantly more space than smaller mammals such as mice. Therefore, there is a need for more
high-throughput models that can replicate genetic-based human skeletal muscle diseases.
1.3 Engineered Tissues as Next-Generation Disease Models

Recently, tissue engineering has been harnessed to develop 2D and 3D in vitro models
that are more high-throughput than animal models while still replicating key architectural
features of native human tissues. Tissue engineered constructs allow for greater control over the
ECM and cell and tissue architecture compared to traditional cell culture. More sophisticated
methods of quantifying the functional phenotype of striated muscle cells and engineered tissues
have also been developed using tissue engineering approaches. However, there is still room for
improvement in current striated muscle tissue engineered constructs to address limitations in
scalability, tunability, and consistency.
20

1.3.1 Existing 2D Engineered Striated Tissue Models
Striated muscle has been
studied in planar, 2D culture in
models ranging from single cells
to cell pairs to multi-cellular
tissues. Most previous single-cell
assays, where single cells are
grown on a plastic petri dish or a
petri dish coated with an ECM
protein, lack the architecture of
native striated muscle cells
91
. However, in more recent work, micropatterning has been used to
create single-cells with the rectangular shape of native striated muscle cells. Micropatterning has
been combined with traction force microscopy to engineer single-cell contractility assays from
which the contractile force of striated muscle cells can be determined. For example, a platform
where single cells and cell pairs could be patterned on a matrix with a tunable protein
composition and elasticity has been previously developed and used with single and paired
neonatal rat ventricular cardiac myocytes
92, 93
and single C2C12 skeletal muscle cells
33
. One
disadvantage of the single-cell skeletal muscle platform is that cells could only be cultured for
five days, and force was quantified for only the first twenty-four hours. While single-cell or cell-
pair studies allow the user to study the effects of myocyte shape on contractility and other
parameters, these are not ideal because they lack the cell-cell interactions seen in multicellular
native muscle tissue.
Figure 3. An image of the ideal disease and drug-testing model:
Organs-on-Chips. Each chip mimics the basic structure, function, and
microenvironment of each organ tissue. Image courtesy of Huh, D.
Trends Cell Biol, 2011
1
.
21

While multicellular cardiac tissues more closely replicate native cardiac tissue than single
cells or cell pairs, the utility of previous multicellular cardiac tissue platforms varies. Some
previously-engineered 2D cardiac tissues lack the architecture of native tissues
94
. Aligned, multi-
cellular 2D cardiac tissues have also been created using micromolded, UV-crosslinked
methacrylated tropoelastin. However, this platform could not be used to quantify contractility
95
.
Muscular thin films (MTFs) are a 2D tissue platform that can also measure contractility. MTFs
consist of micropatterned PDMS
96-98
or micromolded gelatin
99
hydrogels pre-cut into cantilevers.
The cantilevers are parallel to the patterning, and once the tissue has formed, a user peels up the
cantilevers and measures stresses generated by the tissue based on cantilever curvature
97, 100, 101
.
This platform has been successfully used to quantify contractile function in response to drug
exposure
96, 99
. While MTFs mimic striated muscle tissue architecture better than single cells or
cell pairs, this platform is time-consuming to manufacture and requires a large amount of manual
handling. There are many steps involved in manufacturing the MTFs, and manually peeling the
cantilevers can tear the muscle tissues. Another disadvantage of MTFs is that the extracellular
matrix elasticity and protein composition cannot be easily modified. While previous 2D multi-
cellular striated muscle tissues successfully mimic the alignment of native tissues and can
quantify the phenotype of striate muscle tissues, these tissues have limited scalability and control
of disease-relevant microenvironmental factors.
1.3.2 Existing 3D Engineered Tissue Models
While the above-mentioned platforms can replicate the structure of muscle tissue better
than traditional cell culture, they still fall short in replicating the 3D architecture of native
striated muscle, where cells and matrix elements interact in more than one plane. Previous work
shows that cells behave differently in 2D and 3D environments. Skeletal muscle myotubes form
more quickly and more consistently in 3D cell cultures than in standard planar cell culture
102
.
22

Differences in morphology, contractile ability, proliferation, myofibril organization, and
differentiation have been seen between 2D and 3D cell cultures. For example, cardiac cells
grown in 3D have a more round structure and are capable of spontaneous contraction, while
fewer cells grown in 2D exhibit contractions and are more flattened. 2D cardiac cell cultures are
more proliferative, but less differentiated than 3D cardiac cells
103
. Therefore, the creation of 3D
cell cultures would not only have a more biomimetic structure, but also a more biomimetic
function than standard cell culture.
The procedures and results for engineering 3D tissues varies greatly, with some being
more biomimetic and scalable than others. In general, however, 3D engineered tissues remain
inefficient and difficult to reproduce. Some 3D cultures are cell aggregates, which are clumps of
cells grown in suspension
104
or on a matrix
103
. Other cultures involved mixing cells directly into
an extracellular matrix hydrogel
105, 106
. Even though cell aggregates include the cell-cell and cell-
matrix interactions seen in native tissues, cell aggregates do not mimic the highly-aligned layers
formed by native cardiac tissue. Also, the cells in the core of the aggregates will receive less
oxygen, media, and growth factors, making necrosis highly likely. 3D cultures where cells are
mixed directly into a matrix have a structure more similar to native tissues, but they are low-
throughput, and the structure can vary based on the manufacturer of the bundles. It can also be
difficult to reproduce published procedures for making bundles. For bundles that are created by
growing the cell/matrix mixture around posts, it is easy for the bundles to come off of the post
107
.
Various ECM compositions, such as collagen, Matrigel, and fibrin, have been used to engineer
bundles, but some combinations lead to more mature, aligned, and dense tissues than others
108,
109
. 3D muscle bundles engineered by combining cells directly in ECM proteins have been
utilized successfully for drug testing
105
, but the procedure for engineering and testing these
23

bundles must become more high-throughput and consistent to encourage widespread use in
disease modeling and drug-testing.
1.4 Objective and Aims
As described above, both microenvironmental factors and genetic mutations contribute to cardiac
and skeletal muscle diseases. However, current models for disease modeling and drug-testing
(primarily cell culture and animal models) struggle to accurately and controllably replicate the
architecture and microenvironment of striated muscle tissues, leading to bottlenecks in the drug
development process. Thus, our goal is to leverage tissue engineering approaches to engineer
new models of striated muscle tissue that are more biomimetic than conventional cell culture
and more scalable and controllable than animal models. We will focus on engineering platforms
can be used to determine the effects of both microenvironmental factors and genetic mutations
on muscle phenotype in the context of health and disease.
Need: Existing model systems for striated muscle have limited ability to accurately replicate the
architecture, microenvironment, and phenotype of native human tissues in a scalable manner,
hindering the drug development process for both acquired and inherited diseases.
Objective: Engineer scalable and modular model systems of human striated muscle tissue that
can be used to determine the effects of disease-relevant microenvironmental factors and/or
genetic mutations on muscle phenotype.
Aim 1: Engineer a scalable platform to test the effects of ECM and cellular remodeling
on the contractility of 2D engineered cardiac tissues.
Aim 2: Advance the engineering of 3D human skeletal muscle tissues using 3D-printing
for rapid prototyping and tissue clearing for enhanced visualization.
24

Impact: The platforms developed here can be implemented as innovative approaches for disease
modeling and drug screening, including measurements of personalized responses by using
patient-derived cells.

25

2 Aim 1: Engineer a scalable platform to test the effects of ECM and cellular
remodeling on the contractility of 2D engineered cardiac tissues.

Cardiovascular diseases are the leading cause of death in the United States
2
. Furthermore,
cardiotoxicity is a common reason for market withdrawal of drugs
110
. These issues occur
partially because existing model systems cannot accurately recapitulate the complexity of native
myocardium, provide quantitative functional data, and/or achieve reasonable scalability for
medium- to high-throughput disease modeling and drug screening. Myocardium consists of
highly aligned cardiac myocytes and fibroblasts surrounded by a compliant extracellular matrix
(ECM) basement membrane that comprises mostly fibronectin (FN) and laminin (LN).
Importantly, dynamic remodeling of both ECM composition and its mechanical properties,
especially elasticity, have been observed throughout cardiac development and disease
5, 8, 111, 112
,
suggesting they play a prominent role in the physiology of the myocardium. However, animal
models are not effective for controllably identifying how distinct changes to the ECM impact
cardiac function. Animal models are also expensive to maintain, are challenging to scale, and
have limited relevance to humans
45
. Alternatively, in vitro models are less expensive and more
scalable. This approach usually entails culturing cardiac myocytes on highly-simplified surfaces,
such as culture dishes uniformly coated with ECM protein, that poorly recapitulate the
architecture of native myocardium. Furthermore, standard outputs of these models include
morphology, viability, cell beating, and electrophysiology
29, 30
, which are relatively indirect
metrics of cardiac function compared to contractility.
Recently, tissue engineering has been leveraged to develop more sophisticated in vitro
models that mimic key features of native myocardium and integrate built-in assays for
quantifying contractility. For example, stresses generated by single and paired cardiac myocytes
have been measured with traction force microscopy (TFM) by culturing cells on FN-
26

micropatterned hydrogels
113-116
. The advantage of this approach is that both cell shape and the
ECM are relatively tunable. However, these one- and two-cell myocyte constructs lack the
diverse cell demographics and cell-cell interactions seen in native myocardium. Aligned, multi-
cellular cardiac tissues have been engineered on FN-micropatterned PDMS
96-98
or micromolded
gelatin hydrogels
99
pre-cut into mm-scale cantilevers parallel to the orientation of the patterning.
After tissue formation, these cantilevers can be peeled up to measure stresses generated by the
tissue based on cantilever curvature
97, 100, 117
. However, because the cantilevers must be manually
released, this assay is relatively low-throughput. Three-dimensional cardiac tissues have been
generated by encapsulating cells in ECM hydrogels within fabricated molds
118, 119
. These
constructs are likely closer representations of the cell-cell interactions seen in native
myocardium, but the architecture of these constructs is difficult to control and measuring
contractility is cumbersome. These tissues also require relatively large quantities of cells, which
increases cost and limits scalability. Thus, there is a need for a biomimetic myocardial
contractility platform that offers tunability over the ECM and improved scalability compared to
current methods.
To address the limitations of existing models of the myocardium, we engineered a
medium-throughput platform that outputs contractile forces generated by aligned cardiac
myocyte and fibroblast tissues engineered on ECM substrates with tunable mechanical and
biochemical properties. To achieve this, we microcontact printed 200 µm-sided squares of
arrayed 15 µm-wide lines of FN or LN onto polyacrylamide gels with elastic moduli tuned to
mimic healthy (13 kPa) or fibrotic (90 kPa) myocardium
120
. We seeded gels with two different
densities of neonatal rat ventricular cardiac myocytes, which naturally contain cardiac
fibroblasts. The cardiac myocytes and fibroblasts self-assembled into aligned patches of tissue,
27

which we refer to as µMyocardia, with distinct fibroblast:myocyte ratios based on cell seeding
density. Next, we used TFM to quantify peak systolic displacement, cross-sectional force, and
work generated by µMyocardia as a function of ECM elasticity, ECM ligand, and cell
demographics. Our results indicate that peak systolic longitudinal cross-sectional force was
relatively independent of ECM elasticity, ECM ligand, or cell demographics. However, peak
systolic longitudinal displacement and work were significantly lower on more rigid gels. Thus,
ECM elasticity appears to be the most dominant factor regulating contractile output. Because this
platform offers independent control over ECM ligand, ECM elasticity, and cell demographics, it
can be used to engineer customized µMyocardia that mimic unique aspects of healthy and
diseased myocardium. This platform is also relatively scalable compared to existing in vitro
contractility assays and thus can be extended to screening drug toxicity and/or efficacy.
Methods
Photolithography and Soft Lithography
A photolithographic chromium mask containing 200 µm x 200 µm squares of 15-µm
wide lines separated by 2 µm-wide gaps (referred to as 15x2) was designed in AutoCAD
(Autodesk Inc.), similar to previous publications
96, 121
. We chose 200 µm x 200 µm-sized square
tissues because these dimensions were the maximum size that was visible in the field of view of
our inverted fluorescent microscope with a 20x objective, which was essential for resolving
fluorescent beads for TFM measurements.
Silicon wafers were fabricated in a Class 100 Cleanroom using standard photolithography
protocols
122
. Briefly, cleaned silicon wafers were spin coated with hexamethyldisilazane and
then SU-8 2005 negative photoresist (MicroChem). The wafers were exposed to UV light
through the chromium photomask using a mask aligner. The non-cross-linked portions of the
28

wafers were removed by submerging the wafers in propylene glycol monomethyl ether acetate.
Last, the wafers were silanized overnight in a desiccator with a drop of trichloro (1H,1H,2H,2H-
perfluorooctyl) silane. PDMS stamps were created by mixing the base and curing agents of
Sylgard 184 (Dow Corning) at a 10:1 mass ratio and then pouring this mixture over the wafer in
a Petri dish. The PDMS was degassed, cured at 65ºC for four hours, and peeled off the wafer and
cut into square stamps.
Biotinylation of Extracellular Matrix Proteins
FN was biotinylated similar to previously published protocols
114, 115
. Briefly, 200 µg/mL
of biotinylated FN dissolved in sodium carbonate was incubated overnight with Sulfo-NHS-LC-
Biotin (Pierce) in a 100 mM sodium carbonate solution and then dialyzed for 4 hours in
phosphate buffered saline (PBS). LN was biotinylated by dialyzing laminin dissolved in Tris-
HCl and sodium chloride for four hours in a 100 mM sodium carbonate solution. LN was then
incubated overnight with Sulfo-NHS-LC-Biotin (Pierce) in a 100 mM sodium carbonate solution
and dialyzed for four hours in PBS.
Fabrication of Micropatterned Polyacrylamide Gels
Glass coverslips were chemically activated using previously published methods by
immersing the coverslips in 0.5% APTES in 95% ethanol, followed by 70% glutaraldehyde in
water
114, 115
. Polyacrylamide gels with 13 kPa and 90 kPa elastic moduli were fabricated similar
to previously published methods. 40% acrylamide and 2% N,N-methylenebisacrylamide
solutions were mixed in a 5:4 ratio for 13 kPa gels and a 1:1 ratio for 90 kPa gels. For 13 kPa
gels, 240 µL of the acrylamide/bis mixture was combined with 190.1 µL of water. For 90 kPa
gels, 360 µL of the acrylamide/bis mixture was combined with 70.1 µL of water. Each of these
29

mixtures was combined with 60 µL PBS, 0.9 µL tetramethylethylenediamine, and 3 µL
ammonium persulfate. For traction force microscopy (TFM) measurements, 6 µL fluorescent
microbeads were added to the gel solution. For samples used for immunostaining, an extra 6 µL
of water was added to the gel solution. Streptavidin-acrylamide was mixed with the complete gel
solution using a volume ratio of 1:5. 15 µL of the resulting solution was pipetted onto chemically
activated 25 mm coverslips and covered with 18 mm coverslips. The coverslips were left at room
temperature for approximately half an hour until the gel cured, at which point the top 18 mm
coverslips were removed using a razor blade.
To microcontact print the gels, PDMS stamps were sonicated, dried, coated with
biotinylated FN or LN, and incubated for at least one hour at room temperature. The
polyacrylamide gels were placed in a 37ºC incubator for approximately ten minutes to remove
excess moisture from the gel surface, taking care not to over-dry the gels. Then, the stamps were
dried using compressed air and inverted onto the gels such that the patterned side of the stamp
was in contact with the gel. Patterned gels were stored in PBS at 4°C until cell seeding.
Harvest and Culture of Neonatal Rat Ventricular Myocytes
Two-day old Sprague-Dawley rats were euthanized using a protocol approved by the
University of Southern California Institutional Animal Care and Use Committee, similar to
previous protocols
97, 113-115, 117
. Ventricular tissue was isolated from rats, incubated in a 1 mg/mL
trypsin solution for 11-13 hours at 4ºC, and vigorously pipetted in four serial solutions of 1
mg/mL collagenase to dissociate the tissue into a single cell suspension. The cells were strained
and preplated twice to reduce non-myocyte cell populations. The purified cell solution was
seeded onto the micropatterned gels in six-well plates, at a concentration of 5x10
4
cells/cm
2
or
1x10
5
cells/cm
2
per gel. Cells were maintained in M199 media supplemented with 10 mM
30

HEPES, 0.1 mM MEM nonessential amino acids, 20 mM glucose, 2 mM L-glutamine, 1.5 µM
vitamin B-12, 50 U/mL penicillin, and 10% heat-inactivated fetal bovine serum (FBS) for two
days. At this point, FBS was reduced to 2% and media was exchanged every other day.
Immunostaining
After four days of culture, tissues were fixed in 4% paraformaldehyde and permeabilized
with 0.2% Triton X-100. After rinsing samples in PBS, samples were incubated with monoclonal
mouse anti-sarcomeric α-actinin (Sigma, 1:200) primary antibody and rabbit anti-FN primary
antibody (Sigma, 1:200) or rabbit anti-LN primary antibody (Sigma, 1:200) for one hour at room
temperature. The samples were rinsed with PBS and then incubated with Alexa Fluor 546 goat
anti-mouse secondary antibody (Life Technologies, 1:200), 4’,6-diamidino-2-phenylindole
(DAPI, 1:200), Alexa Fluor 488 Phalloidin (Life Technologies, 1:200), and Alexa Fluor 633 goat
anti-rabbit secondary antibody (Life Technologies, 1:200) for one hour at room temperature. The
samples were rinsed in PBS again and then mounted onto glass slides using ProLong Gold
Antifade (Life Technologies).
Quantifying Cell Density and Cytoskeletal Alignment
Stained tissues were imaged using a 20x objective on a Nikon Ti inverted fluorescence
microscope and an Andor Zyla scientific CMOS camera. A cell counter program in ImageJ was
used to count the number of nuclei per µMyocardia for multiple µMyocardia taken from at least
two different harvests for each condition tested. Actin alignment was determined using
previously-described custom MATLAB software created based on a fingerprint detection
algorithm
97
. Briefly, images were converted to binary and skeleton images. Vectors pointing in
the direction of actin fibers were assigned and the maximal eigenvalue of all the vectors was
31

defined as the orientational order parameter, which ranges from zero (for completely isotropic
systems) to one (for completely anisotropic systems).
Quantifying Contractile Parameters with Traction Force Microscopy
Coverslips containing µMyocardia were incubated in Tyrode’s solution (1.8 mM CaCl
2
, 5
mM glucose, 5 mM HEPES, 1 mM MgCl
2
, 5.4 mM KCl, 135 mM NaCl, 0.33 mM NaH
2
PO
4
, pH
7.4) and placed in an incubation chamber set at 37ᴼC on a Nikon Ti inverted fluorescence
microscope. Tissues were electrically stimulated with a field electrode at 13 V and 2 Hz. The
pacing frequency was chosen as 2 Hz because most tissues had spontaneous beat rates greater
than 1 Hz, but could not be consistently captured at 3 Hz. Thus, 2 Hz was the pacing rate with
the most consistent success. A brightfield image of each tissue was taken in order to measure
tissue length, and then three-second videos were captured with a 20x objective and Andor Zyla
scientific CMOS camera at 100 frames/s.
Videos were cropped using ImageJ so that each video started at diastole and comprised
exactly two contractions. In MATLAB, fluorescent bead videos were used to calculate
longitudinal (x-direction) and lateral (y-direction) displacement vectors, u
x,n
and u
y,n
, respectively,
discretized to a grid with n field elements. The dimensions of each field element were 10.4 µm in
the x-direction (∆x) and 10.4 µm in the y-direction (∆y). Displacement was calculated as the
average magnitude of the three highest displacement vectors for the first and second
contractions. Longitudinal (t
x,n
) and lateral (t
y,n
) traction stress vector fields were determined
using Fourier transform traction cytometry, as described in previous publications
113, 123
.
Longitudinal force vectors at each field element (f
x,n
) were calculated by multiplying the traction
stress vectors (t
x,n
) at each field element n by the surface area of the field element ( ∆ x ∆ y):
32

𝑓 𝑥 ,𝑛 = 𝑡 𝑥 ,𝑛 ∆𝑥 ∆𝑦 (1)
Lateral force vectors at each field element (f
y,n
) were calculated in a similar fashion:
𝑓 𝑦 ,𝑛 = 𝑡 𝑦 ,𝑛 ∆𝑥 ∆𝑦 (2)
Average longitudinal force through the cross-section of each µMyocardium (F
x
) was calculated
by multiplying f
x,n
at each field element n by the x-position of the field element (x
n
) and dividing
by the length (L) of the µMyocardium
114
, which was measured for each tissue using ImageJ:
𝐹 𝑥 =
1
𝐿 ∑ 𝑓 𝑥 ,𝑛 𝑥 𝑛 (3)
Average lateral force through the cross-section of the tissue (F
y
) was calculated similarly, using
the y-position of the vector (y
n
):
𝐹 𝑦 =
1
𝐿 ∑ 𝑓 𝑦 ,𝑛 𝑦 𝑛 (4)
Longitudinal work (W
x
) was calculated as follows:
𝑊 𝑥 =
1
2
∑𝑢 𝑥 ,𝑛 𝑓 𝑥 ,𝑛 (5)
Similarly, lateral work (W
y
) was calculated as follows:
𝑊 𝑦 =
1
2
∑𝑢 𝑦 ,𝑛 𝑓 𝑦 ,𝑛 (6)
Peak systolic values for all parameters were defined as the value of that parameter at peak systole
minus its value at diastole. Average time to peak systole was calculated as the difference
between the time-point at the start of contraction and the time-point at maximum W
x
. Normalized
x-displacement was calculated at each x-position by summing the magnitude of all x-
33

displacement vectors at that x-position and normalizing this value by the summed magnitudes of
all x-displacement vectors for that tissue.
Statistical Analysis
All measurements were tested for normality using the Lilliefors Test in MATLAB.
Normally distributed data was analyzed using one-way, two-tailed ANOVA and Tukey’s test for
multiple comparisons in MATLAB with α=0.05. Data that was not normally distributed was
analyzed using a two-tailed Kruskal-Wallis test, and then Bonferroni’s method to test multiple
comparisons in MATLAB with α=0.05. The statistical power was calculated using a post-hoc F-
test and the “ANOVA: fixed effects, omnibus, one-way” option on G*Power
124
. The statistical
test for each condition is noted in the figure legend.
Results
Engineering µMyocardia
Myocardium is densely-packed with aligned cardiac myocytes that are supported by
fibroblasts
125
and a compliant basement membrane consisting primarily of FN and LN
126
. Our
goal was to engineer the minimum functioning unit of myocardium, which we call
µMyocardium, with independent control over ECM ligand and elastic modulus and an integrated
assay for measuring contractility. Previously, we used TFM to measure contractile stresses
generated by single and paired cardiac myocytes cultured on micropatterned polyacrylamide
gels
114, 115
. Here, we expanded this approach to multi-cellular tissues (Figure 6) by fabricating
PDMS stamps with arrays of 200 µm-sided squares. Each square contains 15 µm-wide, 5 µm-
deep lines separated by 2 µm-wide gaps, which is a miniaturized version of a previous pattern
96,
98
. Next, we used these stamps to microcontact print biotinylated FN or LN onto 13 kPa and 90
34

kPa streptavidin-doped polyacrylamide gels to mimic the elasticity of healthy and fibrotic
myocardium
120
, respectively. With this approach, we can mimic the alignment of native
myocardium while independently controlling the biochemical and mechanical properties of the
ECM.
Cellular Architecture of µMyocardia
Next, we seeded our micropatterned gels with primary neonatal rat ventricular myocytes, which
naturally contain a low percentage of fibroblasts that are not completely separated during the
harvest process. To determine the optimal seeding density for creating aligned, confluent
Figure 6: Fabrication of µMyocardium. Tunable polyacrylamide gels containing fluorescent beads and
streptavidin acrylamide were fabricated on chemically activated glass coverslips. PDMS stamps (blue) with 200 µm
x 200 µm squares of 15-µm wide lines separated by 2 µm-wide, 5 µm-deep gaps were coated with biotinylated ECM
proteins (FN or LN, shown in orange), dried, and inverted onto the PA gels, transferring patterned ECM proteins
onto the gels. Patterned gels were then seeded with neonatal rat ventricular myocytes with a natural fraction of
fibroblasts, which self-assembled into aligned µMyocardium. Scale bars, 50 µm.
35

Figure 7: Cell demographics and actin alignment in engineered µMyocardia. (A) Representative images of
micropatterned ECM proteins (i, iv) and µMyocardia engineered using low (ii,v) and high (iii,vi) seeding densities
on FN- and LN-patterned 13 kPa gels. Scale bars, 20 µm; blue, nuclei; red, sarcomeric α-actinin; green, actin. (B)
Total number of cells per µMyocardium (mean ± standard error of the mean; * p< 0.05 for the total cell number and
number of fibroblasts compared to high seeding density for the same ECM elasticity and ligand, ANOVA followed
by Tukey’s test for multiple comparisons). (C) Average actin orientational order parameter for each condition
(mean ± standard error of the mean; * p<0.05 compared to high seeding density for the same ECM elasticity and
ligand, Kruskal-Wallis followed by the Bonferroni method for multiple comparisons).
36

µMyocardia, we seeded constructs with either 5.5x10
4
cells/cm
2
or 1.1x10
5
cells/cm
2
, referred to
as low and high seeding density, respectively. After four days in culture, we stained µMyocardia
for nuclei, actin, and sarcomeric α-actinin (Figure 7A), counted the overall number of nuclei per
µMyocardium, and counted the number of nuclei not located within sarcomeric α-actinin-
positive cells, which we classified as fibroblasts. As expected, low density µMyocardia had a
lower overall number of cells (Figure 7B). Furthermore, high density µMyocardia had
approximately twice as many myocytes compared to low density µMyocardia. This is expected
because myocytes are post-mitotic and therefore final myocyte density should be comparable to
initial myocyte seeding density. Conversely, low density µMyocardia had approximately twice
as many fibroblasts compared to high density µMyocardia (Figure 7B). This is likely because
fibroblasts in both conditions proliferated until they were contact-inhibited, which would occur
earlier in high density µMyocardia. For both low and high density µMyocardia, cell density and
demographics were independent of ECM elasticity and ECM ligand (Figure 7B), indicating that
these factors did not differentially impact cell adhesion, proliferation, or survival.
Because native myocardium is highly aligned, we next determined if our engineered
µMyocardia mimicked this hallmark architectural feature by calculating the orientational order
parameter of actin fibers
97, 117
in multiple µMyocardia for each condition. As shown in Figure
7C, the average actin orientational order parameter for all the conditions was 0.4-0.7, which is a
relatively high level of alignment consistent with values seen previously in aligned tissues
127, 128
.
We found no statistical differences in alignment between the conditions, except for low density
and high density µMyocardia patterned with FN on 13 kPa gels, suggesting that ECM ligand and
elasticity do not considerably impact cytoskeletal alignment.

37

Quantifying Contractile Parameters in Engineered µMyocardia
Because the essential function of myocardium is to contract and pump blood, contractility
is a desirable output for an in vitro model of the myocardium. To facilitate contractility
measurements, we cultured µMyocardia on polyacrylamide gels doped with fluorescent beads
for TFM analysis. After four days in culture, we electrically stimulated µMyocardia at 2 Hz,
inducing tissues to contract and deform the gel. For each µMyocardium, we recorded bead
movement and calculated longitudinal (u
x,n
) and lateral (u
y,n
) displacement vectors and
longitudinal (t
x,n
) and lateral (t
y,n
) traction stress vectors at discretized locations for each
contraction (Figure 8A-C), taking into account the elastic modulus of each gel. We then
Figure 8: Quantifying contractile stresses generated by µMyocardia. (A) Brightfield image of representative
µMyocardium, with tissue edges outlined in white. Corresponding peak systolic displacement (u) (B) and peak
systolic traction stress (t) (B) vectors and colormaps generated by analysis software. Scale bars, 100 µm. (D)
Schematic and calculations for longitudinal force vectors (f
x,n
), longitudinal cross-sectional force (F
x
), and longitudinal
work (W
x
). (E) Longitudinal work plotted over two paced contractions.
38

calculated the longitudinal cross-sectional force (F
x
) and longitudinal work (W
x
) over two
contractions (Figure 8D-E). As shown in the representative heatmaps in Figure 9, displacement
and traction stress vectors at peak systole relative to diastole were predominantly oriented
parallel to the long axis of the tissue, which is expected based on their alignment. Similar to
other studies
114, 115
, displacement vectors were significantly smaller on the more rigid gels
(Figure 9A-B). We also observed that vector maps were similar in magnitude, but more uniform
and symmetrical for the high seeding density compared to the low seeding density (Figure 9A,
C), which is likely due to greater myocyte confluence in the high seeding condition.
Qualitatively, there were no apparent differences in displacement or traction stress vector maps
based on ECM ligand (Figure 9A, D).
To characterize the spatial distribution of force transmission from µMyocardia to the
ECM, we next summed and normalized the magnitudes of peak systolic longitudinal
displacement vectors at each x-position for each µMyocardium. As shown in Figure 10, for all
Figure 9: Displacement and traction stress vector colormaps for multiple µMyocardia. Representative
peak systolic displacement (i) and peak systolic traction stress (ii) vectors and colormaps for indicated ECM
conditions and seeding densities. Scale bars, 100 µm
39

conditions, the average spatial distribution of longitudinal displacement vectors was similar, with
displacement concentrated at the
longitudinal edges of the tissues.
Conversely, in the center of the
tissue, minimal amounts of
displacement were detected.
Thus, in all conditions,
µMyocardia consistently transmitted force predominantly at the longitudinal boundaries of the
tissue.
Comparing Contractile Output of µMyocardia
Next, we calculated time to peak systole, peak systolic longitudinal displacement, and
peak systolic F
x
for multiple µMyocardia for each condition. As shown in Figure 11A, time to
peak systole was similar for all conditions. As suggested by the displacement maps (Figure 9),
peak systolic longitudinal displacement was significantly lower on 90 kPa gels compared to 13
kPa gels (Figure 11B). However, this parameter was not regulated by ECM ligand or cell
density. Interestingly, peak systolic F
x
was similar across all conditions, except for low density
LN 13 kPa differing from high density LN 13 kPa and low density FN 13 kPa (Figure 11C).
Thus, µMyocardia had relatively consistent force production, independent of ECM elasticity,
ECM ligand, or cell demographics. For all conditions, peak systolic displacement, cross-
sectional force, and work in the longitudinal direction were higher than these values in the lateral
direction, as expected due to the uniaxial alignment of the µMyocardia. Thus, peak systolic F
x

generated by µMyocardia is relatively conserved, regardless of ECM ligand, ECM elastic
modulus, and cell seeding density.
Figure 10: Spatial distributions of x-displacement vectors in
µMyocardia. For each condition, the normalized x-displacement was
calculated at each x-position for all µMyocardia, averaged, and plotted.
40

Figure 11: Average longitudinal contractile parameters for µMyocardia. Time to peak systole (A), peak systolic
longitudinal displacement (B), peak systolic longitudinal cross-sectional force F
x
(C), and peak systolic longitudinal
work W
x
(D) for each condition (* p<0.05 compared to 13 kPa, same ECM ligand, same seeding density; † p<0.05
compared to FN, 13 kPa, low seeding density; ‡ p<0.05 compared to LN, 13 kPa, high seeding density; Kruskal-Wallis
followed by the Bonferroni method for multiple comparisons).
41

With each contraction cycle, left ventricular myocardium must both shorten and generate
stress to eject blood into the aorta. Stroke work is defined as the amount of work needed for this
process and is calculated by multiplying stroke volume times aortic pressure, or calculating the
area of a pressure-volume loop over one contraction cycle
129
. In our system, peak systolic
longitudinal W
x
is analogous to stroke work because it considers both displacement and stress
from diastole to systole. Peak systolic W
x
was independent of both ECM ligand and seeding
density (Figure 11D). Thus, µMyocardia generated equivalent levels of work, independent of the
number of cardiac myocytes. However, peak systolic W
x
was significantly higher for
µMyocardia on 13 kPa gels compared to those on 90 kPa gels (Figure 11D). For all conditions,
peak systolic W
x
was higher than W
y
, again due to the alignment of the µMyocardia. Thus,
longitudinal work in µMyocardia was inversely regulated by ECM elastic modulus, but was
independent of both ECM ligand and cell seeding density for the tested ranges.
Identifying Non-Patterned ECM Molecules
Although we micropatterned pure solutions of FN or LN onto gels, ECM molecules are
also synthesized by cells and are present in the media due to our use of serum. Thus, we next
investigated if LN was present in FN-patterned tissue and vice versa by immunostaining for the
ECM molecule that is the counterpart to the micropatterning. As shown in Figure 12, we
detected LN and FN fibers in FN- and LN-patterned tissues, respectively. The LN and FN
immunosignals were primarily localized with fibroblasts instead of myocytes, which is expected.
Thus, µMyocardia synthesized detectable amounts of ECM molecules beyond those initially
micropatterned on the surface, which potentially diminished the effects of the distinct ECM
ligands patterned on the gel.
42

Discussion
To address some of the limitations of
existing in vivo and in vitro models of
myocardium, we engineered a µMyocardium
platform that offers independent control over
ECM elasticity, ECM ligand, and cell
demographics, as well as an integrated
contractility assay. Our results indicate that the
amount of peak systolic W
x
generated by
µMyocardia is regulated by ECM elasticity, but not ECM ligand or the number of myocytes.
These results are important for understanding how distinct cellular and extracellular parameters
independently impact the contractile output of myocardium. Furthermore, our platform is
relatively scalable and therefore can be used for a variety of applications related to disease
modeling and drug screening.
Myocardium is powered by cardiac myocytes that are highly aligned and densely packed
with myofibrils
126
, which maximizes force generation parallel to the long axis of the tissue
97, 101,
130
. To mimic this critical architectural feature, we used photolithography, soft lithography, and
microcontact printing to micropattern arrayed lines of biotinylated FN or LN onto streptavidin-
doped polyacrylamide gels. Although this approach has previously been used for FN
114-116
, our
results show that it is also compatible with LN. On both 13 kPa and 90 kPa gels, micropatterned
FN and LN successfully induced cells to self-assemble into aligned tissues. However, in many
pathological settings, such as the border zone surrounding a myocardial infarction
131
, myocytes
lose their hallmark alignment. Because ECM patterning can be manipulated simply by altering
Figure 12: Expression of non-patterned ECM in
µMyocardia. (A) Representative µMyocardium on
FN-patterned gel and immunostained for LN (green).
(B) Representative µMyocardium on LN-patterned gel
and immunostained for FN (green). Both tissues were
engineered using high seeding density and 13 kPa gels.
Scale bars, 50 µm; blue, nuclei; red, sarcomeric α-
actinin.
43

photomask design, our platform can be easily tuned to engineer cardiac disease models with
physiologically-relevant architectural features. Fine-tuning tissue architecture could also reveal
further insight into relationships between myocyte shape, alignment, and contractility to
complement previous studies in single cells
113, 115
and multicellular, mm-scale tissues
97, 117, 130
.
Furthermore, in this study, we only tested FN and LN individually. However, the basement
membrane of healthy myocardium contains both ligands in defined concentrations and ratios,
which are often altered in many pathological settings
8, 111, 112
. The biochemical characteristics of
physiological and pathological ECM could easily be mimicked in our platform by mixing and
customizing the concentration of FN and LN in the stamping solution. Thus, our approach
enables users to independently dictate tissue alignment, ECM elasticity, and ECM composition,
which is highly beneficial for cardiac disease modeling.
Cardiac myocytes occupy the most volume in myocardium (75%) but comprise only 30-
40% of the myocardium in terms of cell numbers. Fibroblasts comprise the majority of the
remaining cells
125
. Fibroblasts play mostly a supportive role in the myocardium by synthesizing,
depositing, and degrading the ECM
5
. In many pathological settings, such as after a myocardial
infarction, the number of fibroblasts at and/or near the affected area increases due to myocyte
necrosis and fibroblast proliferation
125
. Because we utilized primary neonatal rat ventricular
myocytes in this study, we naturally had a small percentage of fibroblasts present in our
engineered µMyocardia. Since fibroblasts are proliferative, and myocytes are not, the number
and ratio of these two cell types was ultimately regulated by the initial cell seeding density.
Specifically, our high seeding density µMyocardia had 10-15% fibroblasts and low seeding
density µMyocardia had 30-35% fibroblasts. These fibroblast:myocyte ratios are lower than both
healthy and diseased myocardium, but it is challenging to form tissues that maintain electrical
44

coupling between cardiac myocytes in 2D culture if 60-70% of the tissue is fibroblasts. This
inability to truly replicate the 3D architecture and cell positioning of native myocardium is one
limitation of our in vitro system. However, our results still have important implications for
understanding the impact of fibroblasts on myocytes. For example, we found that the initial cell
seeding density, and thus the number of cardiac myocytes, had relatively minimal impact on the
contractile parameters that we measured. This could have a variety of explanations, as detailed
below.
One potential explanation for the independence of contractile output on the number of
myocytes is that fibroblasts had a beneficial effect on the cardiac myocytes. As described above,
we found that the fibroblast:myocyte ratio was higher in the tissues seeded at low density.
Previous studies have shown that increasing the fibroblast:myocyte ratio improves cardiac
myocyte morphology and electrical
132
and contractile
133
phenotype in a variety of engineered
tissue systems. This could be because fibroblasts secrete ECM molecules, which provide
additional mechanical and structural support to cardiac myocytes. Evidence also suggests that
fibroblasts can form functional cell-cell junctions with cardiac myocytes
134, 135
and therefore
likely affect cardiac myocytes in many ways beyond ECM synthesis and remodeling. Together,
our results combined with these previous studies indicate that fibroblasts likely do play an
important role in maintaining and promoting the structural and functional phenotype of cardiac
myocytes in vitro and should not be perceived only as a “contamination” of primary cultures.
These hypotheses can be explored in future studies by more deliberately altering µMyocardia
cellular composition. Our results also suggest that fibroblasts could potentially enhance the
functionality and/or maturity of stem cell-derived cardiac myocytes. Currently, stem cell-derived
cardiac myocytes are typically differentiated, purified, and maintained in monoculture
136, 137
.
45

However, previous studies have shown that co-culturing stem cell-derived cardiac myocytes with
fibroblasts has a beneficial effect on the adhesion and functionality of the myocytes
138-140
, which
is in-line with our current results. Collectively, our study provides further evidence that
fibroblasts are important components to include in both healthy and diseased cardiac tissue
models.
Our previous studies in pairs of cardiac myocytes showed that, at the cell-cell interface,
myocytes predominantly release their focal adhesions and form cohesive cell-cell junctions
114,
116
. Thus, at the cell-cell interface, minimal stresses are transmitted to the ECM, which are the
only stresses that we can measure using TFM. However, we can safely assume that stresses
transmitted at cell-matrix adhesions are appropriately balanced by stresses transmitted to cell-cell
adhesions due to the need for forces to balance
114, 141
. Due to this feature, we previously observed
that traction stresses generated by a single myocyte and a pair of myocytes is relatively
equivalent, if the cell(s) occupy the same surface area and are arranged linearly
114, 115
. Thus, the
surface area of tissue adhesion dominates the amount of stress transmitted to the ECM compared
to the actual number of myocytes. In our µMyocardia, we similarly observed that traction
stresses were concentrated at the longitudinal ends of the tissues over a relatively consistent
surface area, suggesting that cell-cell adhesions were dominant over cell-matrix adhesions in the
interior of the tissue. Furthermore, we observed that contractile parameters were independent of
the number of myocytes. Together, our observations in µMyocardia further support the idea that,
up to a certain point, the geometry and surface area of the tissue, rather than the overall number
of myocytes, dominates the amount of contractile stress transmitted to the ECM, given that ECM
elasticity is kept constant. Overall, our data spanning from single myocyte to myocyte pair to
multi-cellular µMyocardium further demonstrate that the myocardium functions as a mechanical
46

syncytium, where myocytes form robust cell-cell adhesions that seamlessly transmit force across
the tissue.
The ECM is a network of proteins and molecules that have both structural and non-
structural functions. Myocardial ECM is dominated by Type I collagen fibers, which serve
largely structural and stabilizing roles
112
. The basement membrane that directly surrounds
myocytes and links myocytes to collagen fibers consists primarily of the glycoproteins FN and
LN
126
, which are adhesive ligands for integrin receptors
6, 142
. Concurrent remodeling of both
ECM composition and elastic modulus have been reported in many disease settings
111
. For
example, many cardiac diseases are associated with fibrosis and/or scar tissue formation
5
, which
alters ECM composition
8, 143
, including elevated expression of FN
144, 145
, and increases the elastic
modulus of the tissue beyond 50 kPa
120
. With in vivo models, it is challenging, if not impossible,
to truly de-couple ECM composition and elasticity to identify the structural and functional
impact of these parameters. With our in vitro approach, we independently controlled ECM
elasticity and ECM ligand such that we could quantify how these two parameters alter the
contractile output of myocardium. Our results indicate that ECM elasticity has a greater
functional impact on µMyocardia compared to ECM ligand. One explanation for this result is
that, although FN and LN are ligands for distinct integrin receptors
6
, the engagement of different
integrin receptors had minimal impact on contractility. A second explanation is that, because
myocytes and predominantly fibroblasts secrete FN and LN, ultimately both ligands were present
in our constructs, regardless of the identity of the original micropatterned ligand. For example,
our immunostained tissues do show evidence of cell-generated FN in LN-patterned tissues, and
vice versa. Thirdly, our use of serum in the media could have confounded our results because
serum contains many ECM molecules. However, polyacrylamide gels are non-fouling surfaces,
47

which is the reason for using biotin-streptavidin linkages to adhere micropatterned FN and LN to
the surface. As a result, cells did not adhere to any regions of the gel not subjected to
micropatterning, suggesting that ECM molecules from the serum did not deposit on the surface
of the gel. However, we cannot exclude any interactions between soluble ECM molecules in the
serum with the micropatterned ECM molecules or with the cells themselves. Thus, ECM
synthesis by the cells and the presence of ECM molecules in the serum could potentially have
masked any differences due to FN versus LN.
Our results indicate that, in µMyocardia, peak systolic displacement and W
x
were lower
on 90 kPa gels compared to 13 kPa gels, but F
x
was preserved. Interestingly, we observed
slightly different trends in single and paired cardiac myocytes: peak systolic displacement and
various metrics for peak systolic force were lower on 90 kPa gels compared to 13 kPa gels, but
peak systolic work was preserved
114, 115
. Thus, in all cases, peak systolic displacement was
lowest on the more rigid gels. However, in µMyocardia, the difference in displacement between
13 kPa and 90 kPa gels was much greater compared to this difference in single cells and cell
pairs. This ultimately led to a smaller difference in the traction stress and F
x
generated on 13 kPa
and 90 kPa gels, as this value is dependent on both gel elasticity and displacement.
Consequently, peak systolic work in µMyocardia was also much lower on the 90 kPa gels
compared to 13 kPa gels, as this is a function of both displacement and traction stress. Thus,
relationships between displacement, force, and work do not perfectly scale from single cell to
cell pair to multi-cellular tissue. One key difference between µMyocardia and single and paired
myocytes is the presence of fibroblasts, which could potentially add greater resistance to
contraction in µMyocardia. Another important consideration is that we did not measure changes
in diastolic parameters in our µMyocardia, as all measurements reflect changes from diastole to
48

systole. Our reason for excluding these measurements is that measuring baseline diastolic
parameters with TFM requires destroying each µMyocardium (through trypsinization) after
measurements, which would significantly hamper our throughput by limiting each coverslip to
only one µMyocardium. However, there could be differences in baseline diastolic displacement,
force, and/or work that are important for understanding how the parameters we tested potentially
contribute to diseases such as diastolic heart failure
146
. These measurements are an important
topic for follow-up studies.
Many limitations of our study are inherent to in vitro systems. For example, we used
polyacrylamide gels functionalized with ECM proteins such that we could independently
modulate mechanical and biochemical properties of the ECM. However, these gels are highly
synthetic and limit the ability of cells to interact with and remodel the ECM, which is an
important process in native myocardium. Ideally, we would use hydrogels consisting purely of
our ECM molecule of interest, but this is cost-prohibitive. Another limitation is that we only
tested FN and LN, but the basal lamina in native myocardium has many other molecules, such as
Collagen IV
126
. However, our biotin-streptavidin approach should be compatible with testing
other ECM molecules and combinations of ECM molecules in future studies. Future studies
could also investigate lower elastic moduli, such as 1 kPa
114
, to establish how modulating ECM
elasticity and composition impacts contractility in more of a developmental context
147, 148
.
Another limitation is that our constructs are two-dimensional. However, three-
dimensional constructs require large quantities of cells and are less scalable, so although we
compromise physiological relevance, our two-dimensional approach offers many practical
advantages. Furthermore, µMyocardia remain planar throughout the experimental measurements,
require minimal user intervention, and are not endpoint experiments. Thus, our µMyocardia are
49

well-suited to be enclosed in a fluidic device to create a “Myocardium on a Chip” platform that
could be used for extended drug testing or coupling with other “Organs on Chip” platforms
96, 149
.
Our platform is also relatively versatile and should be compatible with measuring contractile
parameters in other types of muscle cells, such as vascular or bronchial smooth muscle.
Our study is also limited by the use of neonatal rat cardiac myocytes, which are non-
human. For future studies, we plan to use this same platform with human induced pluripotent
stem cell (hiPSC)-derived cardiac myocytes sourced from wildtype patients and patients with
inherited cardiomyopathies, similar to previous studies
98, 150, 151
. By combining our µMyocardia
platform with hiPSC-derived cardiac myocytes, we can build robust, human-relevant, patient-
specific platforms to study inherited cardiomyopathies, identify novel therapeutic targets, and
screen drugs with quantitative functional outputs. Importantly, because our platform offers
control over the ECM, we can also determine how diverse forms of ECM remodeling potentially
synergize with genetic mutations in inherited diseases. For example, merosin-deficient
congenital muscular dystrophy is an inherited striated muscle myopathy caused by a mutation in
the LAMA2 gene, which encodes for the LN-α2 chain
152
. Our platform would be especially
powerful for elucidating intracellular and extracellular mechanisms of this disease because we
could independently tune the LN composition of the ECM for both wildtype and LAMA2-
deficient hiPSC-derived cardiac myocytes.
In summary, we have fabricated a scalable platform for engineering and interrogating
customizable µMyocardia, which has revealed novel insight into how remodeling of the ECM
and cellular composition impact the contractile function of myocardium. Our platform has many
applications for functional disease modeling and drug screening, especially when combined with
hiPSC-derived cardiac myocytes.
50

3 Aim 2: Engineer more scalable 3D human skeletal muscle tissues using
3D-printing for rapid prototyping and tissue clearing for enhanced
visualization

Skeletal myopathies encompass a diverse group of disorders that each cause muscle
weakness due to some form of muscle fiber dysfunction. Myopathies can be inherited diseases,
such as Duchenne muscular dystrophy (DMD)
153
or amyotrophic lateral sclerosis
154
, or be
acquired secondary to a variety of extrinsic stressors, including aging
155
, statins
156
, and diabetes
157
. Currently, developing and testing new drugs for inherited or acquired forms of skeletal
myopathies is hindered by the limitations of existing model systems. Animal models are
routinely used for these purposes because they retain the complexity of native muscle. However,
animal models are relatively low-throughput and high-cost and have fundamental genetic,
physiological, and pathophysiological differences from humans. For example, DMD is routinely
modeled in transgenic mdx mice
70, 71
, which show moderate signs of degeneration compared to
DMD patients
158
. A spontaneous golden retriever model of muscular dystrophy is more similar
to DMD
78, 79
, but using these animals is very costly. Furthermore, in DMD and other inherited
myopathies, there are many diverse forms of the disease, each of which is associated with
specific genetic mutations
15, 159
. Thus, it is impractical, or even impossible, to use spontaneous
or transgenic animal models to recapitulate the many unique genotypes and phenotypes of all
inherited human myopathies. Furthermore, the progression of acquired myopathies is known to
be dependent on the genetic background of the patient
160, 161
, which is another feature that
cannot be captured by animal models.
As a simpler alternative to animal models, conventional in vitro techniques are used to
grow skeletal muscle in a dish. This typically entails culturing skeletal myoblasts on polystyrene
plates or dishes and inducing their fusion and differentiation into striated, multi-nucleated
51

myotubes
39, 162, 163
. The major advantage of this technique is that human myoblasts isolated from
muscle biopsies
35, 164
or differentiated from pluripotent stem cells
151, 165
can be grown in vitro to
monitor patient-specific genotypes and phenotypes. This approach is also relatively high-
throughput and low-cost. However, the major limitation is that myotubes form a flat, 2-
dimensional (2-D) layer and remain relatively immature, limiting their physiological relevance.
This immaturity is attributed in part to the artificial environment of 2-D culture, which does not
recapitulate the 3-D, bundle-like architecture of native muscle
13
.
Recently, engineered 3-D skeletal muscle tissues with mm-scale dimensions have
emerged as a medium-throughput, in vitro model that more closely recapitulates the architecture
of native skeletal muscle compared to 2-D monolayers
166, 167
. Similar approaches are used to
engineer 3-D cardiac muscle tissues
168, 169
. In general, the procedure for engineering 3-D skeletal
or cardiac bundles entails pipetting a concentrated mixture of cells and extracellular matrix into a
template that is fabricated by photolithography
170, 171
or milling
105, 172
. Over several days in
culture, the cell/matrix solution compacts and differentiates into a muscle bundle that is anchored
by features built into the template, such as PDMS posts
107, 119
, sutures
171, 173
, or a Velcro or
nylon frame
105, 106
. However, photolithography and milling are relatively time-consuming and
specialized fabrication techniques. Thus, it is challenging to rapidly prototype muscle bundle
templates and screen features such as muscle bundle geometry, which is an important regulator
of myotube distribution and alignment
174
. Furthermore, to visualize the interior of bundles,
researchers currently rely primarily on sectioning and immunostaining
172
. However, sectioning
is relatively tedious and requires a highly-trained user. Even in the best circumstances, obtaining
pristine, well-oriented sections is difficult. Another limitation of sectioning is that the tissue can
only be viewed along the plane in which it was sectioned, limiting the amount of information
52

acquired per sample. This can be an especially large obstacle when engineering muscle bundles
from patient-derived cells, which are often in short supply.
Given these bottlenecks in engineering 3-D muscle bundles, our first goal was to develop
a more rapid technique for fabricating and screening muscle bundles templates. To do this, we
used a consumer-grade 3-D printer to fabricate PDMS templates consisting of rectangular
chambers with a range of dimensions. Within each chamber, we added two sutures to serve as
attachment points for the bundles. We then injected a mixture of C2C12 myoblasts and collagen
I/Matrigel into the template chambers. Over 24 days, we quantified bundle survival and width to
identify the optimal geometries for the chamber. We also verified the contractile ability of select
muscle bundles by recording motion in response to electrical stimulation. Our second goal was to
circumvent the need for sectioning to visualize engineered muscle bundles by implementing
tissue clearing. Tissue clearing is a technique used to increase the optical transparency of tissues
by first cross-linking the tissue into a transparent hydrogel and then extracting the lipids
175, 176
.
The tissue can then be fluorescently tagged and non-obstructively imaged in 3-D
177
. Tissue
clearing has been applied mostly to ex vivo tissues or whole organisms, including brain slices
178
,
mouse organs
179
, and zebrafish
180
. Here, we applied tissue clearing to intact engineered muscle
bundles. Cleared bundles showed similar distributions of myogenic markers compared to tissue
sections. However, unlike tissue sections, we could select the imaging plane on-demand in any
dimension, including longitudinal or cross-sectional views. Thus, we could acquire more visual
information per bundle. Our combined approach of using 3-D printing to rapidly fabricate
muscle bundle templates and using tissue clearing to extract more visual information per bundle
are important technical advances for engineering and interrogating muscle bundles. These
53

techniques can ultimately be applied towards modeling human skeletal myopathies in vitro by
overcoming the limitations of existing animal models and conventional in vitro approaches.
Materials and Methods
3D Printing and Fabrication of Muscle Bundle Templates
Casts were designed using TinkerCAD (Autodesk Inc.) and printed in Acrylonitrile
Butadiene Styrene (ABS) using a Makerbot Replicator 2x 3D Printer (Makerbot Industries). The
cast consisted of a recessed rectangle with width 8.53 mm, length 35.45 mm, and depth 1 or 2
mm. A semicircle with a radius of 3.90 mm was added to one end of the rectangle for ease of
mold removal. A raised rectangle was positioned inside the recessed rectangle with the following
length x width x height dimensions, in mm: 13x2x2, 8x2x2, 8x1x2, and 8x2x1. Cylindrical holes
with a radius of 0.5 mm were placed in parallel with the shorter ends of the raised rectangle, 1
mm from each end. Silk suture (size 3-0, VWR) was strung through the cylindrical holes in the
cast.
Polydimethylsiloxane (PDMS) was produced by mixing the base and curing agents of
Sylgard 184 (Dow Corning) at a 10:1 mass ratio and degassing using a Thinky Mixer AR-100
(Thinky Corporation, Tokyo, Japan). This mixture was poured into the 3D-printed casts and left
in a vacuum desiccator at room temperature overnight for degassing. Then, the 3D-printed casts
were refilled with PDMS 184 to ensure that the casts were completely and evenly filled and left
overnight in a vacuum desiccator at room temperature for a second degassing step. The casts
were then baked for four hours at 65° C to solidify the PDMS. The sutures were then removed
and the PDMS templates were extracted from the casts by gripping the semi-circular tab. The
tabs were then cut off so that the final lengths of the PDMS templates were 19 mm (for 13 mm
long bundle molds) or 14 mm (for 8 mm long bundle molds) to fit onto 25 mm glass coverslips.
54

New sutures were strung through the holes in the templates and tied on the outside of the
template. The bottoms of PDMS templates were then coated with uncured Sylgard 184 and
placed onto a 25 mm glass coverslip as a base. A stainless steel disk weighing approximately
0.45 kg was placed on top of the PDMS molds and the complete constructs were baked for four
hours at 65° C to cure the PDMS and seal the template to the coverslip.
Culture of C2C12 Myoblasts
Mouse C2C12 skeletal myoblasts (ATCC) were seeded and cultured in growth media
containing high glucose Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10%
fetal bovine serum and 1% penicillin-streptomycin. Cells were maintained in an incubator at 37°
C with 5% CO
2
. Growth media was replenished every 2 days. Cells were passaged every 3-4
days. To passage, cells were rinsed with 10 mL PBS and incubated in 10 mL trypsin-EDTA until
most cells had detached. Then, a 1:20 dilution of cells was added to a new flask. Cells between
passages 2-7 were used.
Muscle Bundle Engineering
Complete constructs (PDMS templates attached to glass coverslips) were treated in a
UVO cleaner model 342 (Jelight) for 1 minute for sterilization. Each construct was then placed
in a well of a 6-well plate and incubated in 1% Pluronic 527 (Sigma-Aldrich) for 15 minutes to
prevent bundles from adhering to the PDMS template. C2C12 myoblasts were dissociated using
0.5% trypsin-EDTA (Gibco) and mixed at a concentration of 10e6 cells/mL in 40.2% 1.33
mg/mL collagen I (Corning), 45.3% growth media, 0.3% 0.23 g/L sodium hydroxide (VWR),
and 14.3% undiluted Matrigel (Corning). Then, the cell/matrix solution was pipetted into the
rectangular chambers of the PDMS templates and incubated at 37°C for 30 minutes before
growth media was added. 51 μL, 30 μL, 16 μL, and 16 μL of cell/matrix solution were added to
55

the 13x2x2, 8x2x2, 8x1x2, and 8x2x1 mm molds, respectively. Media was exchanged every 2
days. After four days in culture, bundles were incubated in differentiation media containing high
glucose Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 2% horse serum and
1% penicillin-streptomycin. Bundles were maintained in differentiation media in an incubator at
37° C with 5% CO
2
for 21 additional days. Media was exchanged every two days.
Quantification of Bundle Survival and Width
Bundles were imaged every two days using a Nikon Eclipse TS 100 microscope and
Canon PowerShot S110 camera. Percentage survival was calculated as the number of unbroken
bundles divided by the total number of bundles. Bundle width was calculated by averaging the
width at four evenly-spaced locations on each bundle, measured using ImageJ. This procedure
was repeated for nine bundles total, fabricated as three independent sets of three bundles each.
Electrical Stimulation of Muscle Bundles
After 3 weeks in differentiation media, an engineered muscle bundle, still in the PDMS
template, was incubated in Tyrode’s solution (1.8 mM CaCl
2
, 5 mM glucose, 5 mM HEPES, 1
mM MgCl
2
, 5.4 mM KCl, 135 mM NaCl, 0.33 mM NaH
2
PO
4
, pH 7.4) and placed in an
incubation chamber set at 37ᴼC on a Nikon Ti inverted fluorescence microscope. The bundle was
electrically stimulated with a field electrode at 33 V and 1, 2, and 50 Hz. Brightfield videos of
the center of the contracting bundles were captured with a 10x objective and AndorZyla
scientific CMOS camera at 100 frames/sec. The change in pixel intensity within a region of
interest was quantified and plotted using ImageJ. Change in pixel intensity was normalized as a
ratio of the difference between the average pixel brightness of the current frame and the lowest
average pixel brightness to the maximum range in average pixel brightness.
Tissue Sectioning and Immunostaining
56

Engineered muscle bundles were fixed in 2% paraformaldehyde for 8 hours and then
incubated in 30% sucrose overnight. Samples were then sectioned using a Leica CM3050 S
Research Cryostat at a thickness of 10 µm and placed onto VWR Superfrost Plus Micro Slides.
Sections were encircled using a liquid blocker (Ted Pella). Slides were then rinsed with
phosphate-buffered solution (PBS) and submerged in 10% goat serum (ThermoFisher) and 5%
bovine serum albumin (Hyclone) for 4 hours at room temperature. Sections were incubated with
monoclonal mouse anti-sarcomeric α-actinin primary antibody (Sigma, 1:200) and rabbit anti-
dystrophin primary antibody (AbCam, 1:200) for 24 hours at 4ᴼC. Antibodies were refreshed
after 18 hours of incubation. Sections were rinsed in PBS and then incubated with Alexa Fluor
546 goat anti-mouse secondary antibody (Life Technologies, 1:200), 4’,6-diamidino-2-
phenylindole (DAPI, 1:200), and Alexa Fluor 488 goat anti-rabbit secondary antibody (Life
Technologies, 1:200) for 4-5 hours at room temperature. Sections were rinsed in PBS and
mounted using ProLong Gold Antifade (Life Technologies) and glass coverslips. To image
bundle sections, z-stack confocal images of the middle of the sections were imaged using a 10x
air objective on a Nikon C2 point-scanning confocal microscope system.
Tissue Clearing and Immunostaining
Following published protocols
175
, bundles were rinsed with ice-cold PBS and then
incubated in a solution of 4% acrylamide, 0.05% bisacrylamide, 10% 10X PBS, 4%
paraformaldehyde, 52.5% millipore water, and 0.25% VA-044 thermal initiator for 1 day at 4°
C. The constructs were then vacuum degassed using a nitrogen gas supply and incubated at 37°
C for 4-5 hours to initiate hydrogel cross-linking. Excess hydrogel was then gently removed
using a Kimwipe. The bundles were removed from the PDMS templates and washed twice in a
clearing buffer, prepared by diluting 20% sodium dodecyl sulfate and 1 M boric acid fivefold in
57

distilled water. The bundles (still in clearing buffer) were shaken at 37° C for approximately four
weeks until the bundle was optically clear. Cleared bundles were washed twice in boric acid
buffer (0.2 M, pH 8.5, with 0.1% Triton X-100) for 1 day at 37° C.
Cleared bundles were incubated with monoclonal mouse anti-sarcomeric α-actinin
(Sigma, 1:50) primary antibody and rabbit anti-dystrophin primary antibody (AbCam, 1:50) for 5
days at 37ᴼC. Antibodies were refreshed with 500 µL of antibody/PBS on the second, fourth, and
fifth days of incubation. Bundles were washed in PBS buffer for one day and then incubated with
Alexa Fluor 546 goat anti-mouse secondary antibody (Life Technologies, 1:100), 4’,6-
diamidino-2-phenylindole (DAPI, 1:100), and Alexa Fluor 488 goat anti-rabbit secondary
antibody (Life Technologies, 1:100) for 3 days at 37ᴼC. Then, bundles were rinsed in PBS buffer
for 1 day. Bundles were soaked in FocusClear (CelExplorer Labs Co.) for one hour before
imaging. To image bundles, z-stacks of the middle of the bundles, still in FocusClear, were
imaged using a 10x air objective on a Nikon C2 point-scanning confocal microscope system.
Statistical Analysis
Cox Proportional Hazards Regression was used to compare the survival curves. Bundle
width measurements were tested for normality using the Lilliefors Test in MATLAB. Because
the data was not normally distributed, it was analyzed using a two-tailed Kruskal-Wallis test and
then Bonferroni’s method to test multiple comparisons in MATLAB with α=0.05. The statistical
test for each condition is noted in the figure legend, with additional analyses included in the
Supplementary Tables.
Results
Fabrication of Muscle Bundle Templates
58

As described above, previous approaches to engineer 3-D muscle bundles have relied
primarily on specialized techniques, such as photolithography or milling, to fabricate templates
for molding muscle tissues. Here, we implemented 3-D printing to fabricate PDMS templates for
this purpose (Figure 13A). First, we used TinkerCAD, an open-source software program that
requires minimal technical expertise, to design casts. We then used a consumer-grade 3D printer
to rapidly and cheaply print casts from ABS in approximately half an hour (Figure 13B). Next,
we cured PDMS within the ABS casts (Figure 13C), which ultimately became the templates for
the engineered muscle bundles. Each PDMS template consisted of a rectangular chamber with
one of the following length x width x height dimensions: 13x2x2 mm, 8x2x2 mm, 8x1x2 mm,
and 8x2x1 mm. We chose these dimensions based on previous studies
105, 174
and to test the
effects of template cross-sectional area and length. We also added a thin slab of PDMS to one
side of the template, which acted as a tab to ease removal of the PDMS template from the ABS
cast. This tab was cut off after the PDMS was removed from the cast. Each chamber also had a
pair of holes at each short end. We strung sutures through these holes and tied them outside of
the chamber to serve as anchor points for the muscle bundle (Figure 13D). As the final step, we
attached each PDMS template to a glass coverslip to serve as a base that was optically clear.
Thus, our final PDMS template was cast from 3-D printed ABS and consisted of a rectangular
chamber for molding, anchoring, and imaging engineered muscle bundles.
Characterization of Muscle Bundle Survival Rate and Width
To engineer muscle bundles, we filled the rectangular chambers of the PDMS templates
with a mixture of C2C12 myoblasts, Type I collagen, and Matrigel. Bundles were cultured in
growth media for three days and then switched to differentiation media for three weeks. To
determine the effects of template geometry, we quantified bundle survival rate. After one week
59

Figure 13: Fabrication of Muscle Bundle Templates. (A) Schematic of the overall fabrication process. 3D-printed
ABS casts (B) are used for molding PDMS templates (C). (D) Sutures are inserted into the PDMS template to serve
as anchor points for the muscle bundle. The PDMS template is then attached to a glass coverslip. The rectangular
chamber in the PDMS template is ultimately filled with a cell/matrix solution. The cells are then cultured for 25
days. Scale bars, 5 mm.
of differentiation (Day 11), 4 out of 9 bundles with dimensions of 8x1x2 mm and 8x2x1 mm
survived while 100% (9 out of the 9) of the 13x2x2 mm and 8x2x2 mm bundles survived. These
survival rates plateaued and were stable until Day 24, the end of the experiment (Day 24) (Figure
14B). The differences between the survival rates were not statistically significant
(Supplementary Table S11). Thus, muscle bundle survival rate was mostly dependent on the
cross-sectional area of the template but not the length, as bundles engineered in the 13x2x2 mm
and 8x2x2 mm templates were more stable compared to the 8x1x2 mm and 8x2x1 mm templates.
For all bundles that were intact on the day of measurement, we quantified bundle width
every two days to evaluate compaction and determine relationships between the dimensions of
60

the template and bundle width. By Day 2, the average bundle width for every condition was less
than 1 mm and the average width of the 8x1x2 mm bundles was significantly lower than the
other templates. The width of the 8x1x2 mm bundles was also significantly lower than the 8x2x2
mm bundles and 13x2x2 mm bundles from Days 4-8 and 4-14, respectively. After approximately
two weeks in culture (Day 16), the bundle widths for all templates were not significantly

Figure 14: Engineered muscle bundle survival and compaction. (A) Muscle bundles engineered in 13x2x2 mm,
8x2x2 mm, 8x2x1 mm, and 8x1x2 mm templates on Days 2 and 24. Scale bars, 2 mm. (B) Survival curves for
engineered muscle bundles. (C) Muscle bundle width over time in culture (* indicates p<0.05 between the bundle
widths for 8x2x2 and 8x1x2 molds; ** indicates p<0.05 between the bundle widths for 13x2x2 and 8x1x2 molds; +
indicates p<0.05 between the bundle widths for 8x1x2 and 8x2x1 molds; Kruskal-Wallis followed by the Bonferroni
method for multiple comparisons). Details of statistical analyses are located in Supplementary Tables S1-S2.
61

different and remained constant until the end of the experiment (Day 24). Thus, the cross-
sectional area of the template had an initial impact on bundle width, but these differences were
less apparent at the later time points.
Electrical Stimulation of Engineered Bundles
To ensure that myoblasts fused into contractile myotubes, we electrically stimulated an
8x2x2 mm muscle bundle after 3 weeks of differentiation and recorded movies of bundle
movement. To determine twitch and tetanus responses, we stimulated this bundle at 1 Hz, 2 Hz,
and 50 Hz (Figure 15A). We recorded videos of bundle movement and calculated the average
pixel brightness in a region of interest along the edge of the bundle. As shown in Figure 15B-D,
the bundle moved in sync with the applied frequency of the electrode and had expected twitch
and tetanus responses at low and high frequencies, respectively. These data demonstrate that the
engineered muscle bundles contain functional myotubes that contracted as expected in response
to electrical stimulation.
Comparison of Tissue Sectioning and Tissue Clearing for Imaging Muscle Bundles
To evaluate myotube morphology inside the bundle, we first used the conventional
technique of fixing, cryosectioning, and immunostaining muscle bundles after three weeks of
differentiation. We sectioned bundles in the longitudinal direction and stained sections for nuclei
and the myogenic markers dystrophin and alpha-actinin. We also incubated bundles in only
secondary antibodies as a control to demonstrate that the antibody attachment was binding to
their target proteins (Supplementary Figure 2). In both 8x2x2 mm and 13x2x2 mm bundles, cell
nuclei were detected throughout the bundles, but dystrophin and alpha-actinin were detected
62

primarily at the edges (Figure 16). This suggests that myotube fusion and differentiation
occurred only at the outer edges of the bundle, which has been observed in other studies
109, 181
.

Figure 15: Contractile response of engineered muscle bundles to electrical stimulation. After 25 days in
culture, 8x2x2 mm muscle bundles were electrically stimulated at various frequencies. Videos of the stimulated
muscle bundles were taken, and the change in the average pixel brightness of an inset of the bundle (A) was
quantified for every frame in the video. Bundles were stimulated at 2 Hz (B), 1 Hz (C), and from 2 Hz to 50 Hz,
representing twitch to tetanus (D). Scale bars, 100 mm (i) and 10 mm (ii-iii).
63

Next, we tested if intact muscle bundles could be processed for immunostaining by using
tissue clearing to circumvent some of the technical limitations of cryosectioning. Muscle bundles
differentiated for three weeks were fixed, embedded in a polyacrylamide hydrogel, and
immersed in a clearing solution containing sodium dodecyl sulfate and borate for four weeks
(Figure 17A). Once bundles were optically clear (Figure 17B), we stained intact bundles for
nuclei, dystrophin, and alpha-actinin and collected z-stacks using confocal microscopy. For
comparison, we also followed this same procedure without the clearing solution. First, we
compared slices (Figure 18E), as expected due to the opacity of the tissue. Conversely, in the
cleared bundle, slices from the middle, top, and bottom of the bundle had fluorescent signal that
was equally clear, Figure 16: Images of cyrosectioned and immunostained engineered muscle bundles.
Sections of muscle bundles engineered in 8x2x2 mm and 13x2x2 mm templates, with insets and separated color
channels shown below. Scale bars, 150 µm; blue, nuclei; green, dystrophin; red, sarcomeric α-actinin.
64

visible, and bright (Figure 18B, 18D, and 18F). We also stained a cleared bundle with only
secondary antibodies (Supplementary Figure 3) and detected negligible levels of fluorescence, as
expected.

Figure 17: Schematic of tissue clearing. (A) Tissue clearing entails embedding the muscle bundle in a
polyacrylamide hydrogel and then immersing it in a clearing solution. Once the bundle is optically clear, it is
immunostained, incubated in FocusClear, and imaged with confocal microscopy. (B) Photographs of uncleared and
cleared bundles on top of text. Scale bars, 2 mm.
In comparing images from the cryosectioned bundle to the cleared bundle, it is apparent
that the pattern of fluorescence in the cleared bundle (Figure 18D) matches that from the
sectioned bundle (Figure 16A). Specifically, nuclei are detected throughout the bundle, but most
of the myotubes are clustered along the outer edges of the bundle (Figure 19). However, unlike
the cryosectioned bundle, we could also view the cross-section of the same bundle (Figure 20)
and
65

Figure 18: Comparison of uncleared and cleared muscle bundles. Images of the bottom (A, B), middle (C, D),
and top (E, F) of immunostained uncleared (A, C, E) and cleared (B, D, F) bundles. Scale bars, 150 µm; blue,
nuclei; green, dystrophin; red, sarcomeric α-actinin.
myotubes are clustered along the outer periphery throughout the entire bundle. As a comparison,
the cross-section of the uncleared bundle is blurry and the fluorescent signal gradually weakens
towards the top of the bundle due to the opaqueness of the tissue. Thus, intact bundles that were
cleared and immunostained produced images matching those from bundles processed using
cryosectioning and immunostaining. However, unlike sectioned bundles, cleared bundles
66

Figure 19: Dystrophin and α-actinin in cleared engineered muscle bundles. Confocal images of the middle of
immunostained cleared bundles stained for dystrophin (A) and sarcomeric α-actinin (B). Scale bars, 150 µm; green,
dystrophin; red, sarcomeric α-actinin.
remained intact and thus could be imaged in any plane with confocal microscopy, providing
additional versatility in imaging and morphological analysis.
Discussion
In this study, we report two technical
advances related to engineering and
interrogating engineered muscle bundles. First,
we used 3D printing to fabricate muscle bundle
templates. This is a relatively rapid fabrication
method compared to alternative techniques,
such as photolithography. Using 3-D printed
templates, we systematically determined the
effects of template geometry on bundle survival
and width. Second, we used tissue clearing to enable confocal imaging of intact muscle bundles
Figure 20: Cross-sectional confocal slices of uncleared
and cleared immunostained muscle bundles. Images of
the cross-sections of immunostained uncleared (A) and
cleared (B) bundles. Scale bars, 150 µm; blue, nuclei; green,
dystrophin; red, sarcomeric α-actinin.
67

after immunostaining. This approach is advantageous because it eliminates the need for
cryosectioning and allows users to capture the entire 3-D structure of intact muscle bundles.
Together, these technologies advance the use of engineered muscle bundles for disease modeling
and drug screening.
One of the major challenges for 3-D tissue engineering is the vast number of variables
that influence the final construct. For engineering muscle bundles specifically, previous reports
have demonstrated that matrix protein and elasticity
109, 174, 182
, bundle dimensions
174, 183
, and cell
concentration
181
each affect bundle structure. Optimizing all of these variables can be tedious
and time-consuming. By using 3D-printing, as described here, users can rapidly prototype
muscle bundle templates to accelerate this step of the process and easily tune features related to
template geometry. For example, in this study, we compared 13x2x2 mm, 8x2x2 mm, 8x1x2
mm, and 8x2x1 mm templates to evaluate the effects of the cross-sectional area and length of the
template on bundle survival and width over three weeks in culture. We chose three weeks as our
desired endpoint because myoblasts typically require at least one week in culture to differentiate
into myotubes and additional time improves fusion and differentiation. After 11 days in culture,
only 44.4% (4 out of 9) of the bundles with a reduced cross-sectional area (8x1x2 mm and 8x2x1
mm) survived, while 100% (9 out of 9) of the 8x2x2 mm and 13x2x2 mm bundles survived.
Thus, the bundles in the templates with lower cross-sectional area had a higher chance of
breaking and a lower survival rate. Interestingly, bundle width was sensitive to the cross-
sectional area of the template at early timepoints. However, after two weeks, bundle width was
statistically similar in all conditions. Together, these data indicate that, for the dimensions tested
here, the cross-sectional area of the template affected bundle survival, but did not change the
final width of the muscle bundle. The length of our template did not have an impact on bundle
68

survival or width, as the 13x2x2 mm and 8x2x2 mm bundles were statistically similar in every
metric.
To validate that our engineered muscle bundles were functional, we electrically
stimulated a bundle at 1, 2, and 50 Hz and recorded movies of the bundle response. The
movement of the bundle matched the stimulation frequency of the electrode, including twitch
and tetanus responses. Although this experiment validates that the muscle bundle contained
functional myotubes, this is a relatively qualitative readout. Similar muscle bundle platforms
have been developed with integrated assays for quantifying tissue contractility or
electrophysiology. For example, tissue contractility has been quantified using imaging-based
readouts
119, 184
or force transducers
118, 185, 186
. Electrophysiology has been characterized using
patch clamp
173
or fluorescent imaging of membrane potentials or calcium transients
105, 119, 187
.
The system we developed here could be further enhanced by integrating similar types of
technologies, which are relatively compatible with our design. For example, we could replace
one of the sutures with a conductive material to serve as an electrode to induce contraction. The
suture could also be replaced with a force transducer to quantify contractility. Furthermore,
because the base of our chamber is a glass coverslip, our setup is compatible with fluorescent
live imaging of changes in voltage or calcium in the muscle bundles.
As shown here and in other studies, we observed myotubes concentrated primarily along
the edges of the bundle. This could be due to a lack of diffusion of oxygen and nutrients to the
interior of the bundle. Strategies for introducing perfusion into the engineered muscle bundles,
such as introducing a perfusable wire through the bundle
188
or bioprinting vasculature systems
into the constructs
189
, would help overcome this limitation. The maturation of our constructs
could also be enhanced by incorporating other types of physical cues. For example, studies have
69

shown that intermittent electrical stimulation of skeletal
190
and cardiac
191
muscle bundles
induces structural and functional maturation. Therefore, integrating other physical features, such
as perfusion or electrical stimulation, would likely further improve the structural and functional
phenotypes of our engineered muscle bundles.
Tissue clearing is a relatively new technique that has been used primarily to visualize
small organisms or whole organs, mostly brains
175, 179, 180, 192, 193
. This approach has not been
widely used in the muscle field. Whole mouse muscles have been successfully cleared, but
fluorescent labelling has had mixed success. In one study, researchers were unable to tag cleared
muscle tissue with fluorescent labels
194
. In another study, cleared muscle tissue was successfully
labeled with DAPI and markers for neurons (tyrosine hydroxylase) and blood vessels (CD31),
but the muscle fibers were not specifically labelled
195
. To the best of our knowledge, tissue
clearing has not previously been applied to any engineered muscle constructs. By implementing
tissue clearing, we successfully imaged entire intact bundles stained for DAPI, alpha-actinin, and
dystrophin. The major advantage of this approach compared to cryosectioning is that the entire
intact bundle can be imaged in 3-D instead of only the single plane that was sectioned. Thus,
with cleared and stained bundles, longitudinal sections and cross-sections can be acquired from
the same bundle, increasing the amount of information gained per sample. This can be especially
advantageous when engineering bundles with patient-derived cells, which are precious and
typically challenging to expand in culture. Another advantage of tissue clearing is that the
technique is relatively easy compared to cryosectioning, which requires a highly trained user and
is a labor-intensive process. Although tissue clearing can require several weeks for processing,
multiple bundles can be cleared simultaneously and the technique does not require highly
specialized training. Furthermore, the clearing time can likely be reduced by adjusting
70

parameters, such as the concentrations of bisacrylamide and acrylamide
175
. Thus, our results
indicate that applying tissue clearing to engineered muscle bundles offers a more detailed view
of tissue structure and has technical advantages compared to sectioning.
The types of platforms described here and elsewhere are especially powerful for
engineering patient-specific muscle bundles for personalized disease modeling and drug
screening. These types of approaches are especially needed for the muscular dystrophies due to
the vast genotypic and phenotypic diversity within this class of disease
14
. DMD is the most
common muscular dystrophy and is caused by a variety of mutations in the dystrophin gene
15
.
Similarly, limb-girdle muscular dystrophy has 15 variations, each of which is caused by a
different genetic mutation
16
. To account for these patient-specific features, myoblasts from
patients with muscular dystrophies have been investigated in vitro, although primarily in 2-D
culture
151, 164, 165
. The ability to engineer patient-specific muscle tissues in 3-D and use tissue
clearing to image and evaluate them would be particularly advantageous for modeling muscular
dystrophies because the dystrophin-glycoprotein complex is located around the perimeter of the
muscle fibers. Thus, this structure is more prominent in 3-D fibers compared to 2-D monolayers
and is most visible in cross-sectional views. With better tools for engineering and imaging 3-D
muscle bundles, we can accelerate the search for personalized therapies for muscular dystrophies
and other types of inherited myopathies. These same types of approaches can also be
implemented for characterizing myopathies that are acquired secondary to drugs and other
extrinsic factors.
In summary, we report two technical advances related to engineering and interrogating
muscle bundles for in vitro disease modeling and drug screening. Our approach can be further
improved by increasing the scalability, enhancing myotube distribution and maturity, and
71

integrating functional readouts. Especially by implementing these types of approaches with
patient-derived myoblasts, these types of engineered tissue models overcome several of the
limitations of animal models and conventional cell culture approaches. Thus, these platforms will
have a significant impact on our understanding of inherited and acquired myopathies and will
accelerate the development of new therapies for these conditions.

72

4 Conclusion
Striated muscle diseases, including cardiovascular diseases and muscular dystrophies,
carry a significant burden on the United States. Existing drug testing model systems,
conventional cell culture and animal models, cannot accurately replicate the complexity of
human striated muscle diseases, preventing us from developing treatments and cures for these
diseases. Native striated muscle contains highly aligned muscle cells with sarcomeres that are
surrounded by a chemically and mechanically defined extracellular matrix which can be
remodeled due to disease, changing its properties. Conventional cell culture models involve
growing cells directly on a polystyrene petri dish or on a protein-coated polystyrene petri dish.
These models lack the high alignment and well-defined extracellular matrix seen in native
striated muscle, preventing conventional cell culture from being used to determine the effects of
changes to the extracellular matrix on healthy and diseased striated muscle tissues. While animal
models retain the structure of native striated muscle and can be used to determine the effects of
drugs on healthy and diseased muscle function, the biochemical and mechanical properties of the
animal muscle cannot be individually tuned in order to determine their effects. Also, animal
models are significantly more low-throughput than cell culture models and may have limited
relevance to humans due to genetic differences between species.
Engineered tissues are an intermediary model that can bridge the gaps between these
models, but the scalability, tunability, and consistency of current models needs to be improved
for use in pharmaceutical drug testing. Here, we have engineered 2D and 3D striated muscle
tissues that are more scalable than current engineered tissue models. These platforms could be
further improved in the future by using human cells to improve relevance and by incorporating
new fabrication and sensing technologies to further improve scalability. Eventually, these
73

platforms can be used to model human diseases as well as to create personalized clinical testing
platforms.
4.1 Scalable 2D engineered cardiac tissues demonstrate the effects of disease-
relevant microenvironmental factors on contractile phenotype

Here, we micropatterned 13 kPa and 90 kPa polyacrylamide gels with aligned squares of
fibronectin (FN) or laminin (LN). We seeded gels with two concentrations of primary neonatal
rat ventricular myocytes, which
naturally contain fibroblasts. Cells
assembled into aligned “μMyocardia”
with fibroblast : myocyte ratios
dependent on initial seeding
concentration. Using traction force
microscopy (TFM), we found that the
peak systolic longitudinal cross-
sectional force was similar across conditions, but the peak systolic work was significantly lower
on 90 kPa gels. Therefore, the work generated by the tissues on the stiffer hydrogels, which
mimicked stiffer, post-infarct myocardium, was significantly lower than the work generated on
the less stiff hydrogels, which mimicked healthy myocardium. This mimics how post-infarct
myocardium would be capable of doing less work, and therefore pumping less blood, than
healthy myocardium. Interestingly, the tissues seeded with half as many cardiac myocytes still
generated the same amount of force, suggesting that the fibroblasts may play a role in enhancing
MicroMyocardium function. Finally, ECM composition did not appear to affect cardiac function.
This platform addresses limitations of existing models by creating a more biomimetic
Figure 21: Overview of the fabrication and analysis of 2D
engineered MicroMyocardium. 2D cardiac microtissues, called
MicroMyocardium were engineered on an ECM with a tunable
protein composition and elasticity on which contractility can be
directly measured.
74

model of the myocardium in a tunable microenvironment. Myocardium consists of aligned,
contractile cardiac myocytes interspersed with fibroblasts, and here, we are able to create aligned
tissues containing both cell types. The cellular demographics as well as the biochemical and
mechanical properties of the ECM can also be tuned on this platform, mimicking cellular,
biochemical, and mechanical changes that occur during many cardiac diseases. This platform
retains the ability to easily control cellular architecture and quantify contractility while also
allowing the user to create tunable multicellular tissues in a much more scalable manner than
current engineered tissue platforms such as muscular thin films, which have a long fabrication
time and require human involvement for quantifying contractility. Because our platform provides
independent control over cell–cell and cell–matrix interactions, it has many applications for
cardiac disease modeling.
Using human induced Pluripotent Stem Cell-derived cardiac myocytes (iPSC-CMs) with
this platform would increase its relevance to humans and could lead to the creation of patient-
specific platforms that could be used for more personalized medicine. However, iPSC
differentiation is still fairly inconsistent and many differentiation protocols do not produce
consistent, functionally mature cardiac myocytes. Also, the scalability of iPSC-derived cells is
limited due to the amount of time it takes to obtain cells as well as how few cells can be
obtained
196
. In addition to the challenges that still exist with producing iPSC-CMs, there are also
challenges with using these cells on our platform. The iPSC-CMs do not adhere efficiently to
micropatterned polyacrylamide gels, and so another tunable material, such as PDMS, will need
to be used. Also, the iPSCs do not conform to patterned proteins as effectively as the neonatal rat
cells. This challenge will need to be addressed by using another form of micropatterning or
perhaps using another method for controlling cell architecture, such as micromolding.
75

Micromolding, a method where the desired pattern is molded onto the hydrogel surface
99
, may
provide a stronger alignment cue for iPSC-CMs. While iPSC-CMs are a promising option for
advancing this 2D engineered cardiac tissue platform, there are cell-related and platform-related
challenges that must be surmounted in order to achieve these goals.
Also, even though the platform presented here is more scalable than current platforms, this
platform is still not as high-throughput and scalable as is needed by the pharmaceutical industry.
Micropatterning soft surfaces can be challenging and requires a certain level of human expertise.
Automation of the micropatterning process could help increase the scalability of this platform.
Also, while polyacrylamide gels allowed us to tune the ECM elasticity, cells cannot survive on
polyacrylamide for longer than 5 days. In the native myocardium, the ECM changes significantly
between the first, second, and fourth weeks after an infarct
8
, and having a platform that allows
for longer-term culture as well as more instantaneous control to the ECM elasticity would allow
researchers to mimic post-infarct remodeling in a more long-term manner. Instantaneous control
to the ECM elasticity could be achieved by engineering tissues on a photo-crosslinkable
hydrogel. Last, this platform could be combined in a microfluidic device with an endothelial cell
layer to produce an even more inclusive model of the heart. Incorporating this into a microfluidic
device would also increase the scalability of this platform by reducing the need for manual media
changes.
4.2 The throughput and visualization of 3D engineered skeletal muscle tissues is
improved using 3D-printing and tissue clearing

We utilized 3D-printing and tissue clearing to engineer 3D skeletal muscle bundle tissues
that can be rapidly prototyped and easily visualized. 3D-printed casts were used to create PDMS
molds of various dimensions, and the muscle bundles were created in these PDMS molds.
Bundles were cultured for 25 days to determine the effects of mold dimensions on bundle
76

breakage and bundle width. Decreasing the cross-sectional area of the mold reduced the
percentage survival from 100% to approximately 45%. Bundle width was not significantly
affected by mold geometry after two weeks in culture. When electrically stimulated, bundles
contracted, demonstrating that bundles contained functional myotubes. Last, bundles were
successfully cleared and immunostained, allowing the full structure to be visualized in multiple
planes using a confocal microscope.
Similar to the 2D platform, the throughput and scalability of this platform still need to be
improved in order to be used for pharmaceutical drug testing. While the number of PDMS casts
could easily be increased to make multiple bundle molds at once, the manual threading of the
silk sutures is a large limitation to scale that must be addressed. In the future, the mold design

Figure 22: Overview of fabrication and analysis of 3D engineered skeletal muscle bundles. 3D-printed casts
were used to fabricated PDMS templates, in which muscle bundles of various dimensions were created. Bundles
were characterized based on change in bundle width and bundle longevity as well as bundle contractility in response
to electrical stimulation. Last, in order to enhance visualization of bundles, bundles were cleared and
immunostained.
77

could be modified to allow for the sutures to be cured into the PDMS molds. This system could
also be improved by integrating functional measurements, which could be done by replacing one
of the silk sutures used as a bundle attachment site by a force sensor. Also, bundles could be
engineered with fluorescently-tagged cells in order to measure calcium transients. Last,
engineering bundles with human skeletal myoblasts, healthy and diseased, and incorporating
other cell types, such as satellite cells, would help improve the relevance of this model.
Also, while tissue clearing removed the manual handling required by tissue sectioning,
clearing took four weeks and required a high concentration of antibodies. Optimization could
reduce the amount of time required for clearing as well as the required antibody concentration.
The continued removal of manual fabrication steps, the integration of methods for functional
quantification of muscle bundles, and the optimization of tissue clearing as well as integrating
human skeletal myoblasts would increase the utility of 3D muscle bundles for the drug
development industry.
The work done here could be expanded by engineering bundles with diseased human
myoblasts from patients with diseases such as Duchenne Muscular Dystrophy or Limb-girdle
Muscular Dystrophies and validating these models by testing drugs with known effects on the
engineered tissues. Engineered tissues could also be used to increase our understanding of
skeletal muscle regeneration and to develop clinical therapies or treatments for fibrosis and
muscle damage caused by surgery or injury
197
. Previous in vitro work has demonstrated that
skeletal muscle is sensitive to mechanical loading
198
. For example, mechanically loaded
engineered muscle bundles demonstrate enhanced myoblast alignment, myofiber diameter, and
force generation
199, 200
. Also, clinical evidence and in vivo studies have shown that massage
therapy or other kinds of mechanical manipulation of injured skeletal muscle promotes
78

recovery
201, 202
. Therefore, repeated mechanical stimulations similar to massage therapy could
enhance formation and regeneration of tissues engineered using myoblasts and supporting
satellite cells, producing an in vitro model of skeletal muscle regeneration. Also, skeletal muscle
regeneration and fibrosis could be studied by creating 3D tissues using fibroblasts
203
and
myoblasts, inflicting wounds on these engineered bundles, and then examining the effects of
drugs or cell or growth factor therapies (including satellite cells, IGF-1, HGF, VEGF, and TGF-
beta1)
204, 205
. Therefore, 3D engineered skeletal muscle bundles could be used in the future to
develop treatments or cures for a variety of muscle diseases and injuries.
4.3 Advantages and disadvantages of 2D and 3D systems

The 2D and 3D model systems presented here each have unique advantages and
disadvantages. The 2D system allows the user to control cell architecture more easily and
directly than the 3D system, and functional assays quantifying tissue contractility and calcium
transients are more streamlined on 2D platforms. Also, the user has more control over the
extracellular matrix in the 2D system: the tunability of 3D ECM is limited because certain ECM
compositions may lead to bundle breakage. Despite these advantages, one large disadvantage of
using 2D tissues is that they are inherently less biomimetic than 3D tissues because the cell-cell
interactions, as well as interactions with the extracellular matrix, are limited to one plane. Also,
because the cells on the 2D model system begin to delaminate after one week in culture, this
system may be more appropriate for acute studies, while more long-term, chronic studies may be
more feasible with the 3D system.
Currently, the usage of 3D tissues is limited by the high number of cells required to
engineer tissues. It can be difficult to obtain large numbers of induced pluripotent stem cells or
primary cell samples. However, studies involving multiple cell types or determining the effect of
79

supporting cell populations on engineered tissues (where the cell: supporting cell ratio is varied)
may be more easily done using 3D tissues, as here the ratios of cells can be directly varied and
controlled when the tissues are being made. Also, some lines of induced pluripotent stem cell-
derived cardiac myocytes seem to easily differentiate when embedded in a protein matrix
206
, and
do not follow micropatterned cues as well, and so 3D systems may be a more feasible platform
for these cells. Functional assessment of 3D tissues still requires some advances in sensor or
imaging technology to allow us to quantify the force or calcium transients in a 3D object.
Therefore, a 2D or 3D model may be better suited depending on the research question being
asked or cells being used, and it is up to the researcher to determine which model system is more
appropriate.

Table 1: Comparison of 2D and 3D engineered tissues. 2D and 3D tissue platforms have different advantages and
disadvantages that researchers must consider when picking a platform.
4.4 Organs-on-Chips in the Future
We have presented 2D and 3D engineered tissue model systems as an intermediary that
can bridge the gaps between model systems that are currently available (in vitro cell culture and
animal models). Because in vitro cell culture is the most high-throughput method, this could be
used as the initial method for screening drugs, and then once the number of possible compounds
has been reduced, animal models and human organs-on-chips could be used simultaneously to
determine whether drug candidates negatively affect tissue function. This would provide users
2D Engineered Tissues 3D Engineered Tissues
 Controllable ECM X Limited control over the ECM
X Cell-cell interactions limited to one plane  More biomimetic cell-cell interactions
 Scalable  Scalable
 Low cell number required X High cell number required
 Can easily quantify function X Cannot easily quantify function
 Human-relevant  Human-relevant
X Less physiological  More physiological
X Usable only for short-term culture  Usable for long-term culture
80

with a more biomimetic model on which to determine if drugs negatively affect human striated
muscle tissue function before clinical trials.
In the future, the use of organs-on-chips could be expanded to not only supplement other
model systems, but also to provide a platform for personalized medicine. Induced pluripotent
stem cells could be used to engineer a personalized organ-on-chip directly from a patient’s cells,
and then compounds could be tested to determine their individual effect on a patient. This would
allow us to treat acquired diseases more effectively and genetic diseases, of which there can be
thousands of variations, more efficiently. The process of fabricating and maintaining organs on
chips must be advanced further to reduce human involvement for successful commercialization
and widespread use by people who are not tissue engineering experts. Also, the efficiency of
obtaining primary cells and creating induced pluriportent stem cell-derived cells must be
improved.

Figure 23: The place of organ-on-a-chip as model systems. Organ-on-chip models currently can be used to
supplement in vitro research and translational models, such as animals, prior to clinical trials. In the future, the
organ-on-a-chip platforms can be advanced for use in personalized clinical research.
81

References
1. D. Huh, G. A. Hamilton and D. E. Ingber, Trends Cell Biol, 2011, 21, 745-754.
2. D. Mozaffarian, E. J. Benjamin, A. S. Go, D. K. Arnett, M. J. Blaha, M. Cushman, S. R. Das, S. de Ferranti, J.
P. Després, H. J. Fullerton, V. J. Howard, M. D. Huffman, C. R. Isasi, M. C. Jiménez, S. E. Judd, B. M. Kissela,
J. H. Lichtman, L. D. Lisabeth, S. Liu, R. H. Mackey, D. J. Magid, D. K. McGuire, E. R. Mohler, C. S. Moy, P.
Muntner, M. E. Mussolino, K. Nasir, R. W. Neumar, G. Nichol, L. Palaniappan, D. K. Pandey, M. J. Reeves, C.
J. Rodriguez, W. Rosamond, P. D. Sorlie, J. Stein, A. Towfighi, T. N. Turan, S. S. Virani, D. Woo, R. W. Yeh,
M. B. Turner, W. G. Members, A. H. A. S. Committee and S. S. Subcommittee, Circulation, 2016, 133, e38-
360.
3. C. F. Spurney, Muscle Nerve, 2011, 44, 8-19.
4. C. A. Walker, The Journal of Thoracic and Cardiovascular Surgery, 1999, 118, 375-382.
5. D. Fan, A. Takawale, J. Lee and Z. Kassiri, Fibrogenesis Tissue Repair, 2012, 5, 15.
6. R. S. Ross and T. K. Borg, Circ Res, 2001, 88, 1112-1119.
7. D. Vanhoutte, M. Schellings, Y. Pinto and S. Heymans, Cardiovasc Res, 2006, 69, 604-613.
8. K. E. Sullivan, K. P. Quinn, K. M. Tang, I. Georgakoudi and L. D. Black, Stem Cell Res Ther, 2014, 5, 14.
9. J. L. Zellner, F. G. Spinale, D. M. Eble, K. W. Hewett and F. A. Crawford, Circ Res, 1991, 69, 590-600.
10. F. G. Spinale, M. Tomita, J. L. Zellner, J. C. Cook, F. A. Crawford and M. R. Zile, Am J Physiol, 1991, 261,
H308-318.
11. F. G. Spinale, D. A. Hendrick, F. A. Crawford, A. C. Smith, Y. Hamada and B. A. Carabello, Am J Physiol,
1990, 259, H218-229.
12. P. W. Armstrong, T. P. Stopps, S. E. Ford and A. J. de Bold, Circulation, 1986, 74, 1075-1084.
13. A. R. Gillies and R. L. Lieber, Muscle Nerve, 2011, 44, 318-331.
14. A. Y. Manzur and F. Muntoni, Postgrad Med J, 2009, 85, 622-630.
15. D. J. Blake, A. Weir, S. E. Newey and K. E. Davies, Physiol Rev, 2002, 82, 291-329.
16. V. Straub and K. Bushby, Semin Pediatr Neurol, 2006, 13, 104-114.
17. K. H. Holt and K. P. Campbell, J Biol Chem, 1998, 273, 34667-34670.
18. Y. Huang, A. de Morrée, A. van Remoortere, K. Bushby, R. R. Frants, J. T. den Dunnen and S. M. van der
Maarel, Hum Mol Genet, 2008, 17, 1855-1866.
19. S. Baghdiguian, M. Martin, I. Richard, F. Pons, C. Astier, N. Bourg, R. T. Hay, R. Chemaly, G. Halaby, J.
Loiselet, L. V. Anderson, A. Lopez de Munain, M. Fardeau, P. Mangeat, J. S. Beckmann and G. Lefranc, Nat
Med, 1999, 5, 503-511.
20. K. J. Schimmel, D. J. Richel, R. B. van den Brink and H. J. Guchelaar, Cancer Treat Rev, 2004, 30, 181-191.
21. A. Albini, G. Pennesi, F. Donatelli, R. Cammarota, S. De Flora and D. M. Noonan, J Natl Cancer Inst, 2010,
102, 14-25.
22. P. Menna, E. Salvatorelli and G. Minotti, Chem Res Toxicol, 2008, 21, 978-989.
23. R. A. Wilke, D. W. Lin, D. M. Roden, P. B. Watkins, D. Flockhart, I. Zineh, K. M. Giacomini and R. M. Krauss,
Nat Rev Drug Discov, 2007, 6, 904-916.
24. N. Yokoo, S. Baba, S. Kaichi, A. Niwa, T. Mima, H. Doi, S. Yamanaka, T. Nakahata and T. Heike, Biochem
Biophys Res Commun, 2009, 387, 482-488.
25. A. Bongso, C. Y. Fong, S. C. Ng and S. Ratnam, Hum Reprod, 1994, 9, 2110-2117.
26. A. J. Boyle, I. K. McNiece and J. M. Hare, Methods Mol Biol, 2010, 660, 65-84.
27. L. da Silva Meirelles, P. C. Chagastelles and N. B. Nardi, J Cell Sci, 2006, 119, 2204-2213.
28. A. I. Caplan, J Orthop Res, 1991, 9, 641-650.
29. R. T. Dorr, K. A. Bozak, N. G. Shipp, M. Hendrix, D. S. Alberts and F. Ahmann, Cancer research, 1988, 48,
5222-5227.
30. V. Shirhatti, M. George, R. Chenery and G. Krishna, Toxicol Appl Pharmacol, 1986, 84, 173-191.
31. J. Kyselovic, J. J. Leddy, A. Ray, J. Wigle and B. S. Tuana, J Biol Chem, 1994, 269, 21770-21777.
32. D. D. Allen, R. Caviedes, A. M. Cárdenas, T. Shimahara, J. Segura-Aguilar and P. A. Caviedes, Drug Dev Ind
Pharm, 2005, 31, 757-768.
33. B. Li, M. Lin, Y. Tang, B. Wang and J. H. Wang, J Biomech, 2008, 41, 3349-3353.
34. R. Caviedes, J. L. Liberona, J. Hidalgo, S. Tascon, K. Salas and E. Jaimovich, Biochim Biophys Acta, 1992,
1134, 247-255.
82

35. C. S. Cheng, Y. El-Abd, K. Bui, Y. E. Hyun, R. H. Hughes, W. E. Kraus and G. A. Truskey, Am J Physiol Cell
Physiol, 2014, 306, C385-395.
36. T. Braun, E. Bober, G. Buschhausen-Denker, S. Kohtz, K. H. Grzeschik, H. H. Arnold and S. Kotz, EMBO J,
1989, 8, 3617-3625.
37. S. J. Rhodes and S. F. Konieczny, Genes Dev, 1989, 3, 2050-2061.
38. D. Montarras, J. Chelly, E. Bober, H. Arnold, M. O. Ott, F. Gros and C. Pinset, New Biol, 1991, 3, 592-600.
39. P. Y. Wang, H. Thissen and W. B. Tsai, Biotechnology and Bioengineering, 2012, 109, 2104-2115.
40. M. T. Lam, S. Sim, X. Zhu and S. Takayama, Biomaterials, 2006, 27, 4340-4347.
41. Y. Sun, R. Duffy, A. Lee and A. W. Feinberg, Acta Biomater, 2013, 9, 7885-7894.
42. R. M. Duffy, Y. Sun and A. W. Feinberg, Ann Biomed Eng, 2016.
43. J. Owens, K. Moreira and G. Bain, In Vitro Cell Dev Biol Anim, 2013, 49, 695-705.
44. R. Caviedes, P. Caviedes, J. L. Liberona and E. Jaimovich, Muscle Nerve, 1994, 17, 1021-1028.
45. V. G. Desai, E. H. Herman, C. L. Moland, W. S. Branham, S. M. Lewis, K. J. Davis, N. I. George, T. Lee, S. Kerr
and J. C. Fuscoe, Toxicol Appl Pharmacol, 2013, 266, 109-121.
46. P. M. Alderton, J. Gross and M. D. Green, Cancer Res, 1992, 52, 194-201.
47. E. H. Herman, A. Rahman, V. J. Ferrans, J. A. Vick and P. S. Schein, Cancer Res, 1983, 43, 5427-5432.
48. R. Fritsche, T. Schwerte and B. Pelster, Am J Physiol Regul Integr Comp Physiol, 2000, 279, R2200-2207.
49. L. I. Zon and R. T. Peterson, Nat Rev Drug Discov, 2005, 4, 35-44.
50. D. J. Hsieh and C. F. Liao, Br J Pharmacol, 2002, 137, 782-792.
51. P. Mulder, V. Richard, P. Compagnon, J. P. Henry, F. Lallemand, J. P. Clozel, R. Koen, B. Macé and C.
Thuillez, J Am Coll Cardiol, 1997, 29, 416-421.
52. U. Langheinrich, Bioessays, 2003, 25, 904-912.
53. R. L. HAMLIN, Am J Physiol, 1960, 198, 537-542.
54. R. L. Hamlin, Pharmacol Ther, 2007, 113, 276-295.
55. D. J. Hearse and F. J. Sutherland, Pharmacol Res, 2000, 41, 597-603.
56. K. Kobayashi, T. Izawa, M. Kuwamura and J. Yamate, J Toxicol Pathol, 2012, 25, 135-147.
57. U. Roostalu and U. Strähle, Dev Cell, 2012, 22, 515-529.
58. D. Montarras, A. L'honoré and M. Buckingham, FEBS J, 2013, 280, 4036-4050.
59. C. Seger, M. Hargrave, X. Wang, R. J. Chai, S. Elworthy and P. W. Ingham, Dev Dyn, 2011, 240, 2440-2451.
60. T. E. Hall, R. J. Bryson-Richardson, S. Berger, A. S. Jacoby, N. J. Cole, G. E. Hollway, J. Berger and P. D.
Currie, Proc Natl Acad Sci U S A, 2007, 104, 7092-7097.
61. M. F. Goody, M. W. Kelly, C. J. Reynolds, A. Khalil, B. D. Crawford and C. A. Henry, PLoS Biol, 2012, 10,
e1001409.
62. D. I. Bassett, R. J. Bryson-Richardson, D. F. Daggett, P. Gautier, D. G. Keenan and P. D. Currie,
Development, 2003, 130, 5851-5860.
63. G. Kawahara, J. A. Karpf, J. A. Myers, M. S. Alexander, J. R. Guyon and L. M. Kunkel, Proc Natl Acad Sci U S
A, 2011, 108, 5331-5336.
64. G. Kawahara, M. J. Gasperini, J. A. Myers, J. J. Widrick, A. Eran, P. R. Serafini, M. S. Alexander, M. T.
Pletcher, C. A. Morris and L. M. Kunkel, Hum Mol Genet, 2014, 23, 1869-1878.
65. H. Hirata, L. Saint-Amant, J. Waterbury, W. Cui, W. Zhou, Q. Li, D. Goldman, M. Granato and J. Y. Kuwada,
Development, 2004, 131, 5457-5468.
66. E. M. Gibbs, E. J. Horstick and J. J. Dowling, Febs j, 2013, 280, 4187-4197.
67. W. R. Telfer, A. S. Busta, C. G. Bonnemann, E. L. Feldman and J. J. Dowling, Hum Mol Genet, 2010, 19,
2433-2444.
68. H. Hirata, T. Watanabe, J. Hatakeyama, S. M. Sprague, L. Saint-Amant, A. Nagashima, W. W. Cui, W. Zhou
and J. Y. Kuwada, Development, 2007, 134, 2771-2781.
69. J. J. Dowling, S. Arbogast, J. Hur, D. D. Nelson, A. McEvoy, T. Waugh, I. Marty, J. Lunardi, S. V. Brooks, J. Y.
Kuwada and A. Ferreiro, Brain, 2012, 135, 1115-1127.
70. L. J. Harcourt, J. D. Schertzer, J. G. Ryall and G. S. Lynch, Neuromuscul Disord, 2007, 17, 47-55.
71. E. Gussoni, Y. Soneoka, C. D. Strickland, E. A. Buzney, M. K. Khan, A. F. Flint, L. M. Kunkel and R. C.
Mulligan, Nature, 1999, 401, 390-394.
72. G. Bulfield, W. G. Siller, P. A. Wight and K. J. Moore, Proc Natl Acad Sci U S A, 1984, 81, 1189-1192.
73. G. R. Coulton, J. E. Morgan, T. A. Partridge and J. C. Sloper, Neuropathol Appl Neurobiol, 1988, 14, 53-70.
83

74. C. Pastoret and A. Sebille, J Muscle Res Cell Motil, 1993, 14, 423-431.
75. E. E. Dupont-Versteegden and R. J. McCarter, Muscle Nerve, 1992, 15, 1105-1110.
76. S. Pierno, B. Nico, R. Burdi, A. Liantonio, M. P. Didonna, V. Cippone, B. Fraysse, J. F. Rolland, D. Mangieri,
F. Andreetta, P. Ferro, C. Camerino, A. Zallone, P. Confalonieri and A. De Luca, Neuropathol Appl
Neurobiol, 2007, 33, 344-359.
77. H. G. Radley, M. J. Davies and M. D. Grounds, Neuromuscul Disord, 2008, 18, 227-238.
78. J. N. Kornegay, S. M. Tuler, D. M. Miller and D. C. Levesque, Muscle & nerve, 1988, 11, 1056-1064.
79. N. J. H. Sharp, 1992, 13, 115–121.
80. J. N. Kornegay, N. J. Sharp, R. O. Schueler and C. W. Betts, Lab Anim Sci, 1994, 44, 331-333.
81. J. N. Kornegay, N. J. Sharp, D. J. Bogan, S. D. Van Camp, J. R. Metcalf and R. O. Schueler, J Neurol Sci, 1994,
123, 100-107.
82. J. N. Kornegay, D. J. Bogan, J. R. Bogan, M. K. Childers, D. D. Cundiff, G. F. Petroski and R. O. Schueler, J
Neurol Sci, 1999, 166, 115-121.
83. R. H. T. Edwards, D. A. Jones, D. J. Newham and S. J. Chapman, Lancet, 1984, 323.
84. J. H. Postlethwait, Y. L. Yan, M. A. Gates, S. Horne, A. Amores, A. Brownlie, A. Donovan, E. S. Egan, A.
Force, Z. Gong, C. Goutel, A. Fritz, R. Kelsh, E. Knapik, E. Liao, B. Paw, D. Ransom, A. Singer, M. Thomson,
T. S. Abduljabbar, P. Yelick, D. Beier, J. S. Joly, D. Larhammar, F. Rosa, M. Westerfield, L. I. Zon, S. L.
Johnson and W. S. Talbot, Nat Genet, 1998, 18, 345-349.
85. A. Amores, A. Force, Y. L. Yan, L. Joly, C. Amemiya, A. Fritz, R. K. Ho, J. Langeland, V. Prince, Y. L. Wang, M.
Westerfield, M. Ekker and J. H. Postlethwait, Science, 1998, 282, 1711-1714.
86. A. L. Siegel, D. B. Gurevich and P. D. Currie, Febs j, 2013, 280, 4074-4088.
87. G. Hollway and P. Currie, Birth Defects Res C Embryo Today, 2005, 75, 172-179.
88. M. Buckingham and S. D. Vincent, Curr Opin Genet Dev, 2009, 19, 444-453.
89. J. Dangain and G. Vrbova, Muscle Nerve, 1984, 7, 700-704.
90. Y. Tanabe, K. Esaki and T. Nomura, Acta Neuropathol, 1986, 69, 91-95.
91. J. R. Gershlak, J. I. Resnikoff, K. E. Sullivan, C. Williams, R. M. Wang and L. D. Black, Biochem Biophys Res
Commun, 2013, 439, 161-166.
92. M. L. McCain, H. Yuan, F. S. Pasqualini, P. H. Campbell and K. K. Parker, Am J Physiol Heart Circ Physiol,
2014.
93. M. L. McCain, H. Lee, Y. Aratyn-Schaus, A. G. Kléber and K. K. Parker, Proc Natl Acad Sci U S A, 2012, 109,
9881-9886.
94. L. Wang, C. Xu, Y. Zhu, Y. Yu, N. Sun, X. Zhang, K. Feng and J. Qin, Lab Chip, 2015, 15, 4283-4290.
95. N. Annabi, K. Tsang, S. M. Mithieux, M. Nikkhah, A. Ameri, A. Khademhosseini and A. S. Weiss, Adv Funct
Mater, 2013, 23.
96. A. Agarwal, J. A. Goss, A. Cho, M. L. McCain and K. K. Parker, Lab Chip, 2013, 13, 3599-3608.
97. A. Grosberg, P. W. Alford, M. L. McCain and K. K. Parker, Lab on a chip, 2011, 11, 4165-4173.
98. G. Wang, M. L. McCain, L. Yang, A. He, F. S. Pasqualini, A. Agarwal, H. Yuan, D. Jiang, D. Zhang, L. Zangi, J.
Geva, A. E. Roberts, Q. Ma, J. Ding, J. Chen, D. Z. Wang, K. Li, J. Wang, R. J. Wanders, W. Kulik, F. M. Vaz,
M. A. Laflamme, C. E. Murry, K. R. Chien, R. I. Kelley, G. M. Church, K. K. Parker and W. T. Pu, Nature
medicine, 2014, 20, 616-623.
99. M. L. McCain, A. Agarwal, H. W. Nesmith, A. P. Nesmith and K. K. Parker, Biomaterials, 2014, 35, 5462-
5471.
100. P. W. Alford, A. W. Feinberg, S. P. Sheehy and K. K. Parker, Biomaterials, 2010, 31, 3613-3621.
101. A. W. Feinberg, P. W. Alford, H. Jin, C. M. Ripplinger, A. A. Werdich, S. P. Sheehy, A. Grosberg and K. K.
Parker, Biomaterials, 2012, 33, 5732-5741.
102. A. Khodabukus and K. Baar, Tissue Engineering: Part C, 2009.
103. C. P. Soares, V. Midlej, M. E. Weschollek de Oliveira, M. Benchimol, M. L. costa and C. Mermelstein, PLoS
One, 2012, 7, e38147.
104. V. Z. Beachley, M. T. Wolf, K. Sadtler, S. S. Manda, H. Jacobs, M. Blatchley, J. S. Bader, A. Pandey, D.
Pardoll and J. H. Elisseeff, Nat Methods, 2015, 12, 1197-1204.
105. L. Madden, M. Juhas, W. E. Kraus, G. A. Truskey and N. Bursac, eLife, 2015, 4, e04885.
106. D. W. van der Schaft, A. C. van Spreeuwel, K. J. Boonen, M. L. Langelaan, C. V. Bouten and F. P. Baaijens, J
Vis Exp, 2013, e4267.
84

107. M. S. Sakar, D. Neal, T. Boudou, M. A. Borochin, Y. Li, R. Weiss, R. D. Kamm, C. S. Chen and H. H. Asada,
Lab Chip, 2012, 12, 4976-4985.
108. S. Battista, D. Guarnieri, C. Borselli, S. Zeppetelli, A. Borzacchiello, L. Mayol, D. Gerbasio, D. R. Keene, L.
Ambrosio and P. A. Netti, Biomaterials, 2005, 26, 6194-6207.
109. S. Hinds, W. Bian, R. G. Dennis and N. Bursac, Biomaterials, 2011, 32, 3575-3583.
110. N. Ferri, P. Siegl, A. Corsini, J. Herrmann, A. Lerman and R. Benghozi, Pharmacology & therapeutics, 2013,
138, 470-484.
111. M. L. McCain and K. K. Parker, Pflugers Archiv : European journal of physiology, 2011, 462, 89-104.
112. S. Jane-Lise, S. Corda, C. Chassagne and L. Rappaport, Heart failure reviews, 2000, 5, 239-250.
113. P. L. Kuo, H. Lee, M. A. Bray, N. A. Geisse, Y. T. Huang, W. J. Adams, S. P. Sheehy and K. K. Parker, The
American journal of pathology, 2012, 181, 2030-2037.
114. M. L. McCain, H. Lee, Y. Aratyn-Schaus, A. G. Kleber and K. K. Parker, Proceedings of the National Academy
of Sciences of the United States of America, 2012, 109, 9881-9886.
115. M. L. McCain, H. Yuan, F. S. Pasqualini, P. H. Campbell and K. K. Parker, American journal of physiology.
Heart and circulatory physiology, 2014, 306, H1525-1539.
116. Y. Aratyn-Schaus, F. S. Pasqualini, H. Yuan, M. L. McCain, G. J. Ye, S. P. Sheehy, P. H. Campbell and K. K.
Parker, The Journal of cell biology, 2016, 212, 389-397.
117. A. W. Feinberg, P. W. Alford, H. Jin, C. M. Ripplinger, A. A. Werdich, S. P. Sheehy, A. Grosberg and K. K.
Parker, Biomaterials, 2012, 33, 5732--5741.
118. W. Bian, C. P. Jackman and N. Bursac, Biofabrication, 2014, 6, 024109.
119. T. Boudou, W. R. Legant, A. Mu, M. A. Borochin, N. Thavandiran, M. Radisic, P. W. Zandstra, J. A. Epstein,
K. B. Margulies and C. S. Chen, Tissue engineering. Part A, 2012, 18, 910-919.
120. M. F. Berry, A. J. Engler, Y. J. Woo, T. J. Pirolli, L. T. Bish, V. Jayasankar, K. J. Morine, T. J. Gardner, D. E.
Discher and H. L. Sweeney, American journal of physiology. Heart and circulatory physiology, 2006, 290,
H2196-2203.
121. G. Wang, M. L. McCain, L. Yang, A. He, F. S. Pasqualini, A. Agarwal, H. Yuan, D. Jiang, D. Zhang, L. Zangi, J.
Geva, A. E. Roberts, Q. Ma, J. Ding, J. Chen, D. Z. Wang, K. Li, J. Wang, R. J. Wanders, W. Kulik, F. M. Vaz,
M. A. Laflamme, C. E. Murry, K. R. Chien, R. I. Kelley, G. M. Church, K. K. Parker and W. T. Pu, Nature
medicine, 2014.
122. D. Qin, Y. Xia and G. M. Whitesides, Nature protocols, 2010, 5, 491-502.
123. J. P. Butler, I. M. Tolic-Norrelykke, B. Fabry and J. J. Fredberg, American journal of physiology. Cell
physiology, 2002, 282, C595-605.
124. F. Faul, E. Erdfelder, A. G. Lang and A. Buchner, Behav Res Methods, 2007, 39, 175-191.
125. P. Camelliti, T. K. Borg and P. Kohl, Cardiovascular research, 2005, 65, 40-51.
126. C. A. Walker and F. G. Spinale, The Journal of thoracic and cardiovascular surgery, 1999, 118, 375-382.
127. A. Agarwal, Y. Farouz, A. P. Nesmith, L. F. Deravi, M. L. McCain and K. K. Parker, Advanced Functional
Materials, 2013, 23, 3738--3746.
128. M. L. McCain, S. P. Sheehy, A. Grosberg, J. A. Goss and K. K. Parker, Proceedings of the National Academy
of Sciences of the United States of America, 2013, 110, 9770-9775.
129. W. Schramm, J Clin Monit Comput, 2010, 24, 213-217.
130. M. B. Knight, N. K. Drew, L. A. McCarthy and A. Grosberg, Biophysical journal, 2016, 110, 1615-1624.
131. T. Matsushita, M. Oyamada, K. Fujimoto, Y. Yasuda, S. Masuda, Y. Wada, T. Oka and T. Takamatsu,
Circulation research, 1999, 85, 1046-1055.
132. B. R. Desroches, P. Zhang, B. R. Choi, M. E. King, A. E. Maldonado, W. Li, A. Rago, G. Liu, N. Nath, K. M.
Hartmann, B. Yang, G. Koren, J. R. Morgan and U. Mende, American journal of physiology. Heart and
circulatory physiology, 2012, 302, H2031-2042.
133. H. Saini, A. Navaei, A. Van Putten and M. Nikkhah, Advanced healthcare materials, 2015, 4, 1961-1971.
134. G. Gaudesius, M. Miragoli, S. P. Thomas and S. Rohr, Circulation research, 2003, 93, 421-428.
135. P. Kohl and R. G. Gourdie, Journal of molecular and cellular cardiology, 2014, 70C, 37-46.
136. P. W. Burridge, E. Matsa, P. Shukla, Z. C. Lin, J. M. Churko, A. D. Ebert, F. Lan, S. Diecke, B. Huber, N. M.
Mordwinkin, J. R. Plews, O. J. Abilez, B. Cui, J. D. Gold and J. C. Wu, Nature methods, 2014, 11, 855-860.
137. X. Lian, J. Zhang, S. M. Azarin, K. Zhu, L. B. Hazeltine, X. Bao, C. Hsiao, T. J. Kamp and S. P. Palecek, Nature
protocols, 2013, 8, 162--175.
85

138. I. C. Parrag, P. W. Zandstra and K. A. Woodhouse, Biotechnology and bioengineering, 2012, 109, 813-822.
139. K. Pfannkuche, S. Neuss, F. Pillekamp, L. P. Frenzel, W. Attia, T. Hannes, J. Salber, M. Hoss, M. Zenke, B. K.
Fleischmann, J. Hescheler and T. Saric, Stem cells and development, 2010, 19, 1589-1599.
140. J. Xi, M. Khalil, D. Spitkovsky, T. Hannes, K. Pfannkuche, W. Bloch, T. Saric, K. Brockmeier, J. Hescheler and
F. Pillekamp, Stem cells and development, 2011, 20, 821-830.
141. V. Maruthamuthu, B. Sabass, U. S. Schwarz and M. L. Gardel, Proceedings of the National Academy of
Sciences of the United States of America, 2011, 108, 4708-4713.
142. M. Rienks, A. P. Papageorgiou, N. G. Frangogiannis and S. Heymans, Circulation research, 2014, 114, 872--
888.
143. P. Oliviero, C. Chassagne, N. Salichon, A. Corbier, G. Hamon, F. Marotte, D. Charlemagne, L. Rappaport and
J. L. Samuel, Cardiovascular research, 2000, 46, 346-355.
144. J. L. Samuel, A. Barrieux, S. Dufour, I. Dubus, F. Contard, V. Koteliansky, F. Farhadian, F. Marotte, J. P.
Thiery and L. Rappaport, The Journal of clinical investigation, 1991, 88, 1737-1746.
145. M. M. Ulrich, A. M. Janssen, M. J. Daemen, L. Rappaport, J. L. Samuel, F. Contard, J. F. Smits and J. P.
Cleutjens, Journal of molecular and cellular cardiology, 1997, 29, 2533-2543.
146. G. P. Aurigemma, M. R. Zile and W. H. Gaasch, Circulation, 2006, 113, 296-304.
147. A. J. Engler, C. Carag-Krieger, C. P. Johnson, M. Raab, H. Y. Tang, D. W. Speicher, J. W. Sanger, J. M. Sanger
and D. E. Discher, Journal of cell science, 2008, 121, 3794-3802.
148. S. Majkut, T. Idema, J. Swift, C. Krieger, A. Liu and D. E. Discher, Curr Biol, 2013, 23, 2434-2439.
149. S. N. Bhatia and D. E. Ingber, Nature biotechnology, 2014, 32, 760--772.
150. J. T. Hinson, A. Chopra, N. Nafissi, W. J. Polacheck, C. C. Benson, S. Swist, J. Gorham, L. Yang, S. Schafer, C.
C. Sheng, A. Haghighi, J. Homsy, N. Hubner, G. Church, S. A. Cook, W. A. Linke, C. S. Chen, J. G. Seidman
and C. E. Seidman, Science, 2015, 349, 982-986.
151. C. S. Young, M. R. Hicks, N. V. Ermolova, H. Nakano, M. Jan, S. Younesi, S. Karumbayaram, C. Kumagai-
Cresse, D. Wang, J. A. Zack, D. B. Kohn, A. Nakano, S. F. Nelson, M. C. Miceli, M. J. Spencer and A. D. Pyle,
Cell stem cell, 2016, 18, 533-540.
152. M. Durbeej, Curr Top Membr, 2015, 76, 31-60.
153. A. E. Emery, Lancet, 2002, 359, 687-695.
154. M. C. Kiernan, S. Vucic, B. C. Cheah, M. R. Turner, A. Eisen, O. Hardiman, J. R. Burrell and M. C. Zoing,
Lancet, 2011, 377, 942-955.
155. T. N. Kim and K. M. Choi, Journal of bone metabolism, 2013, 20, 1-10.
156. S. Nazir, S. Lohani, N. Tachamo, D. Poudel and A. Donato, J Clin Rheumatol, 2017, 23, 149-154.
157. C. M. F. Monaco, C. G. R. Perry and T. J. Hawke, Current opinion in neurology, 2017, 30, 545-552.
158. G. B. Banks and J. S. Chamberlain, Current topics in developmental biology, 2008, 84, 431-453.
159. A. E. Renton, A. Chio and B. J. Traynor, Nat Neurosci, 2014, 17, 17-23.
160. A. Ghatak, O. Faheem and P. D. Thompson, Atherosclerosis, 2010, 210, 337-343.
161. T. P. O'Hanlon and F. W. Miller, Curr Rheumatol Rep, 2009, 11, 287-294.
162. V. Chaturvedi, D. E. Dye, B. F. Kinnear, T. H. van Kuppevelt, M. D. Grounds and D. R. Coombe, PloS one,
2015, 10, e0127675.
163. A. Bettadapur, G. C. Suh, N. A. Geisse, E. R. Wang, C. Hua, H. A. Huber, A. A. Viscio, J. Y. Kim, J. B. Strickland
and M. L. McCain, Scientific reports, 2016, 6, 28855.
164. A. P. Nesmith, M. A. Wagner, F. S. Pasqualini, B. B. O'Connor, M. J. Pincus, P. R. August and K. K. Parker,
The Journal of cell biology, 2016, 215, 47-56.
165. G. C. Kendall, E. I. Mokhonova, M. Moran, N. E. Sejbuk, D. W. Wang, O. Silva, R. T. Wang, L. Martinez, Q. L.
Lu, R. Damoiseaux, M. J. Spencer, S. F. Nelson and M. C. Miceli, Sci Transl Med, 2012, 4, 164ra160.
166. A. Khodabukus, N. Prabhu, J. Wang and N. Bursac, Advanced healthcare materials, 2018, 7, e1701498.
167. G. A. Truskey, Lab on a chip, 2018, 18, 3061-3073.
168. G. Conant, B. F. L. Lai, R. X. Z. Lu, A. Korolj, E. Y. Wang and M. Radisic, Stem cell reviews, 2017, 13, 335-346.
169. N. R. Ariyasinghe, D. M. Lyra-Leite and M. L. McCain, American journal of physiology. Heart and circulatory
physiology, 2018.
170. W. R. Legant, A. Pathak, M. T. Yang, V. S. Deshpande, R. M. McMeeking and C. S. Chen, Proc Natl Acad Sci
U S A, 2009, 106, 10097-10102.
86

171. Y. Xiao, B. Zhang, H. Liu, J. W. Miklas, M. Gagliardi, A. Pahnke, N. Thavandiran, Y. Sun, C. Simmons, G.
Keller and M. Radisic, Lab on a chip, 2014, 14, 869-882.
172. V. Y. Sidorov, P. C. Samson, T. N. Sidorova, J. M. Davidson, C. C. Lim and J. P. Wikswo, Acta biomaterialia,
2017, 48, 68-78.
173. V. Y. Sidorov, P. C. Samson, T. N. Sidorova, J. M. Davidson, C. C. Lim and J. P. Wikswo, Acta Biomaterialia,
2017, 48, 68-78.
174. M. Costantini, S. Testa, E. Fornetti, A. Barbetta, M. Trombetta, S. M. Cannata, C. Gargioli and A. Rainer,
Frontiers in bioengineering and biotechnology, 2017, 5, 22.
175. R. Tomer, L. Ye, B. Hsueh and K. Deisseroth, Nat Protoc, 2014, 9, 1682-1697.
176. K. Chung, J. Wallace, S. Y. Kim, S. Kalyanasundaram, A. S. Andalman, T. J. Davidson, J. J. Mirzabekov, K. A.
Zalocusky, J. Mattis, A. K. Denisin, S. Pak, H. Bernstein, C. Ramakrishnan, L. Grosenick, V. Gradinaru and K.
Deisseroth, Nature, 2013, 497, 332-337.
177. P. Ariel, Int J Biochem Cell Biol, 2017, 84, 35-39.
178. K. H. Jensen and R. W. Berg, Sci Rep, 2016, 6, 32674.
179. J. R. Epp, Y. Niibori, H. L. Liz Hsiang, V. Mercaldo, K. Deisseroth, S. A. Josselyn and P. W. Frankland, eNeuro,
2015, 2.
180. M. R. Cronan, A. F. Rosenberg, S. H. Oehlers, J. W. Saelens, D. M. Sisk, K. L. Jurcic Smith, S. Lee and D. M.
Tobin, Dis Model Mech, 2015, 8, 1643-1650.
181. C. Rhim, D. A. Lowell, M. C. Reedy, D. H. Slentz, S. J. Zhang, W. E. Kraus and G. A. Truskey, Muscle Nerve,
2007, 36, 71-80.
182. Y. J. Choi, T. G. Kim, J. Jeong, H. G. Yi, J. W. Park, W. Hwang and D. W. Cho, Adv Healthc Mater, 2016, 5,
2636-2645.
183. A. J. Capel, R. P. Rimington, J. W. Fleming, D. J. Player, L. A. Baker, M. C. Turner, J. M. Jones, N. R. W.
Martin, R. A. Ferguson, V. C. Mudera and M. P. Lewis, Front Bioeng Biotechnol, 2019, 7, 20.
184. X. Zhang, S. Hong, R. Yen, M. Kondash, C. E. Fernandez and G. A. Truskey, Lab Chip, 2018, 18, 2787-2796.
185. C. E. Rupert and K. L. K. Coulombe, Stem Cells Int, 2017, 2017, 7648409.
186. A. K. Schroer, M. S. Shotwell, V. Y. Sidorov, J. P. Wikswo and W. D. Merryman, Acta Biomaterialia, 2017,
48, 79-87.
187. B. Liau, N. Christoforou, K. W. Leong and N. Bursac, Biomaterials, 2011, 32, 9180-9187.
188. M. Radisic, L. Yang, J. Boublik, R. J. Cohen, R. Langer, L. E. Freed and G. Vunjak-Novakovic, Am J Physiol
Heart Circ Physiol, 2004, 286, H507-516.
189. Y. S. Zhang, A. Arneri, S. Bersini, S. R. Shin, K. Zhu, Z. Goli-Malekabadi, J. Aleman, C. Colosi, F. Busignani, V.
Dell'Erba, C. Bishop, T. Shupe, D. Demarchi, M. Moretti, M. Rasponi, M. R. Dokmeci, A. Atala and A.
Khademhosseini, Biomaterials, 2016, 110, 45-59.
190. A. Khodabukus, L. Madden, N. K. Prabhu, T. R. Koves, C. P. Jackman, D. M. Muoio and N. Bursac,
Biomaterials, 2019, 198, 259-269.
191. K. Ronaldson-Bouchard, S. P. Ma, K. Yeager, T. Chen, L. Song, D. Sirabella, K. Morikawa, D. Teles, M.
Yazawa and G. Vunjak-Novakovic, Nature, 2018, 556, 239-243.
192. K. Ando, Q. Laborde, A. Lazar, D. Godefroy, I. Youssef, M. Amar, A. Pooler, M. C. Potier, B. Delatour and C.
Duyckaerts, Acta Neuropathol, 2014, 128, 457-459.
193. S. Sindhwani, A. M. Syed, S. Wilhelm and W. C. Chan, Bioconjug Chem, 2017, 28, 253-259.
194. A. Milgroom and E. Ralston, Cell Biol Int, 2016, 40, 478-483.
195. W. L. Zhang, S. H. Liu, W. C. Zhang, W. Hu, M. Jiang, A. Tamadon and Y. Feng, Cell J, 2018, 20, 132-137.
196. I. Batalov and A. W. Feinberg, Biomark Insights, 2015, 10, 71-76.
197. N. J. Turner and S. F. Badylak, Cell Tissue Res, 2012, 347, 759-774.
198. T. A. Hornberger, D. D. Armstrong, T. J. Koh, T. J. Burkholder and K. A. Esser, Am J Physiol Cell Physiol,
2005, 288, C185-194.
199. H. H. Vandenburgh and P. Karlisch, In Vitro Cell Dev Biol, 1989, 25, 607-616.
200. d. G. Moon, G. Christ, J. D. Stitzel, A. Atala and J. J. Yoo, Tissue Eng Part A, 2008, 14, 473-482.
201. P. Weerapong, P. A. Hume and G. S. Kolt, Sports Med, 2005, 35, 235-256.
202. C. A. Cezar, E. T. Roche, H. H. Vandenburgh, G. N. Duda, C. J. Walsh and D. J. Mooney, Proc Natl Acad Sci U
S A, 2016, 113, 1534-1539.
87

203. M. S. Sakar, J. Eyckmans, R. Pieters, D. Eberli, B. J. Nelson and C. S. Chen, Nature Communications, 2016, 7,
11036.
204. R. W. Ten Broek, S. Grefte and J. W. Von den Hoff, J Cell Physiol, 2010, 224, 7-16.
205. T. Laumonier and J. Menetrey, J Exp Orthop, 2016, 3, 15.
206. J. Zhang, M. Klos, G. F. Wilson, A. M. Herman, X. Lian, K. K. Raval, M. R. Barron, L. Hou, A. G. Soerens, J. Yu,
S. P. Palecek, G. E. Lyons, J. A. Thomson, T. J. Herron, J. Jalife and T. J. Kamp, Circ Res, 2012, 111, 1125-
1136.

88

Supplemental 1: Aim 2

Supplementary Figure S1. Average lateral contractile parameters for µMyocardia. Peak
systolic lateral displacement (A) and peak systolic lateral cross-sectional force F
y
(B) Kruskal-
Wallis followed by the Bonferroni method for multiple comparisons). Details of statistical
analyses are located in Supplementary Tables S9-S10.

89

Supplementary Table S1. Statistical analysis for total nuclei in µMyocardia.
All data was normally distributed, as determined by the Lilliefors test. The p-value for the
ANOVA test was 1.2012e-19, F=20.81, and there were 7 degrees of freedom. Multiple
comparisons were performed using Tukey’s test, with p values for statistical differences
indicated in the table below. The statistical power was 99.9%.

Comparison p-value
FN, 13 kPa, high vs. LN, 13 kPa, high NS
FN, 13 kPa, high vs. FN, 90 kPa, high NS
FN, 13 kPa, high vs. LN, 90 kPa, high NS
FN, 13 kPa, high vs. FN, 13 kPa, low p<0.05
FN, 13 kPa, high vs. LN, 13 kPa, low p<0.05
FN, 13 kPa, high vs. FN, 90 kPa, low p<0.05
FN, 13 kPa, high vs. LN, 90 kPa, low p<0.05
LN, 13 kPa, high vs. FN, 90 kPa, high NS
LN, 13 kPa, high vs. LN, 90 kPa, high NS
LN, 13 kPa, high vs. FN, 13 kPa, low p<0.05
LN, 13 kPa, high vs. LN, 13 kPa, low p<0.05
LN, 13 kPa, high vs. FN, 90 kPa, low p<0.05
LN, 13 kPa, high vs. LN, 90 kPa, low NS
FN, 90 kPa, high vs. LN, 90 kPa, high NS
FN, 90 kPa, high vs. FN, 13 kPa, low p<0.05
FN, 90 kPa, high vs. LN, 13 kPa, low p<0.05
FN, 90 kPa, high vs. FN, 90 kPa, low p<0.05
FN, 90 kPa, high vs. LN, 90 kPa, low p<0.05
LN, 90 kPa, high vs. FN, 13 kPa, low p<0.05
LN, 90 kPa, high vs. LN, 13 kPa, low p<0.05
LN, 90 kPa, high vs. FN, 90 kPa, low p<0.05
LN, 90 kPa, high vs. LN, 90 kPa, low p<0.05
FN, 13 kPa, low vs. LN, 13 kPa, low NS
FN, 13 kPa, low vs. FN, 90 kPa, low NS
FN, 13 kPa, low vs. LN, 90 kPa, low NS
LN, 13 kPa, low vs. FN, 90 kPa, low NS
LN, 13 kPa, low vs. LN, 90 kPa, low NS
FN, 90 kPa, low vs. LN, 90 kPa, low NS

90

Supplementary Table S2. Statistical analysis for number of cardiac myocytes in
µMyocardia.
All data was normally distributed, as determined by the Lilliefors test. The p-value for the
ANOVA test was 1.93599e-30, F=37.4, and there were 7 degrees of freedom. Multiple
comparisons were performed using Tukey’s test, with p values for statistical differences
indicated in the table below. The statistical power was 100%.

Comparison p-value
FN, 13 kPa, high vs. LN, 13 kPa, high NS
FN, 13 kPa, high vs. FN, 90 kPa, high NS
FN, 13 kPa, high vs. LN, 90 kPa, high NS
FN, 13 kPa, high vs. FN, 13 kPa, low p<0.05
FN, 13 kPa, high vs. LN, 13 kPa, low p<0.05
FN, 13 kPa, high vs. FN, 90 kPa, low p<0.05
FN, 13 kPa, high vs. LN, 90 kPa, low p<0.05
LN, 13 kPa, high vs. FN, 90 kPa, high NS
LN, 13 kPa, high vs. LN, 90 kPa, high NS
LN, 13 kPa, high vs. FN, 13 kPa, low p<0.05
LN, 13 kPa, high vs. LN, 13 kPa, low p<0.05
LN, 13 kPa, high vs. FN, 90 kPa, low p<0.05
LN, 13 kPa, high vs. LN, 90 kPa, low p<0.05
FN, 90 kPa, high vs. LN, 90 kPa, high NS
FN, 90 kPa, high vs. FN, 13 kPa, low p<0.05
FN, 90 kPa, high vs. LN, 13 kPa, low p<0.05
FN, 90 kPa, high vs. FN, 90 kPa, low p<0.05
FN, 90 kPa, high vs. LN, 90 kPa, low p<0.05
LN, 90 kPa, high vs. FN, 13 kPa, low p<0.05
LN, 90 kPa, high vs. LN, 13 kPa, low p<0.05
LN, 90 kPa, high vs. FN, 90 kPa, low p<0.05
LN, 90 kPa, high vs. LN, 90 kPa, low p<0.05
FN, 13 kPa, low vs. LN, 13 kPa, low NS
FN, 13 kPa, low vs. FN, 90 kPa, low NS
FN, 13 kPa, low vs. LN, 90 kPa, low NS
LN, 13 kPa, low vs. FN, 90 kPa, low NS
LN, 13 kPa, low vs. LN, 90 kPa, low NS
FN, 90 kPa, low vs. LN, 90 kPa, low NS

91

Supplementary Table S3. Statistical analysis for number of fibroblasts in µMyocardia.
All data was not normally distributed, as determined by the Lilliefors test. The p-value for the
Kruskal-Wallis test was 1.2242e-13, the chi-square value was 75.34, and there were 7 degrees of
freedom. Multiple comparisons were performed using the Bonferroni method, with p values for
statistical differences indicated in the table below. The statistical power was 100%.

Comparison p-value
FN, 13 kPa, high vs. LN, 13 kPa, high NS
FN, 13 kPa, high vs. FN, 90 kPa, high NS
FN, 13 kPa, high vs. LN, 90 kPa, high NS
FN, 13 kPa, high vs. FN, 13 kPa, low p<0.05
FN, 13 kPa, high vs. LN, 13 kPa, low p<0.05
FN, 13 kPa, high vs. FN, 90 kPa, low p<0.05
FN, 13 kPa, high vs. LN, 90 kPa, low p<0.05
LN, 13 kPa, high vs. FN, 90 kPa, high NS
LN, 13 kPa, high vs. LN, 90 kPa, high NS
LN, 13 kPa, high vs. FN, 13 kPa, low p<0.05
LN, 13 kPa, high vs. LN, 13 kPa, low p<0.05
LN, 13 kPa, high vs. FN, 90 kPa, low p<0.05
LN, 13 kPa, high vs. LN, 90 kPa, low p<0.05
FN, 90 kPa, high vs. LN, 90 kPa, high NS
FN, 90 kPa, high vs. FN, 13 kPa, low p<0.05
FN, 90 kPa, high vs. LN, 13 kPa, low p<0.05
FN, 90 kPa, high vs. FN, 90 kPa, low p<0.05
FN, 90 kPa, high vs. LN, 90 kPa, low p<0.05
LN, 90 kPa, high vs. FN, 13 kPa, low p<0.05
LN, 90 kPa, high vs. LN, 13 kPa, low p<0.05
LN, 90 kPa, high vs. FN, 90 kPa, low p<0.05
LN, 90 kPa, high vs. LN, 90 kPa, low p<0.05
FN, 13 kPa, low vs. LN, 13 kPa, low NS
FN, 13 kPa, low vs. FN, 90 kPa, low NS
FN, 13 kPa, low vs. LN, 90 kPa, low NS
LN, 13 kPa, low vs. FN, 90 kPa, low NS
LN, 13 kPa, low vs. LN, 90 kPa, low NS
FN, 90 kPa, low vs. LN, 90 kPa, low NS

92

Supplementary Table S4. Statistical analysis for actin alignment in µMyocardia.
All data was not normally distributed, as determined by the Lilliefors test. The p-value for the
Kruskal-Wallis test was 0.0002, the chi-square value was 28.28, and there were 7 degrees of
freedom. Multiple comparisons were performed using the Bonferroni method, with p values for
statistical differences indicated in the table below. The statistical power is 98.46%.

Comparison p-value
FN, 13 kPa, high vs. LN, 13 kPa, high NS
FN, 13 kPa, high vs. FN, 90 kPa, high NS
FN, 13 kPa, high vs. LN, 90 kPa, high p<0.05
FN, 13 kPa, high vs. FN, 13 kPa, low p<0.05
FN, 13 kPa, high vs. LN, 13 kPa, low NS
FN, 13 kPa, high vs. FN, 90 kPa, low NS
FN, 13 kPa, high vs. LN, 90 kPa, low p<0.05
LN, 13 kPa, high vs. FN, 90 kPa, high NS
LN, 13 kPa, high vs. LN, 90 kPa, high NS
LN, 13 kPa, high vs. FN, 13 kPa, low NS
LN, 13 kPa, high vs. LN, 13 kPa, low NS
LN, 13 kPa, high vs. FN, 90 kPa, low NS
LN, 13 kPa, high vs. LN, 90 kPa, low NS
FN, 90 kPa, high vs. LN, 90 kPa, high NS
FN, 90 kPa, high vs. FN, 13 kPa, low p<0.05
FN, 90 kPa, high vs. LN, 13 kPa, low NS
FN, 90 kPa, high vs. FN, 90 kPa, low NS
FN, 90 kPa, high vs. LN, 90 kPa, low NS
LN, 90 kPa, high vs. FN, 13 kPa, low NS
LN, 90 kPa, high vs. LN, 13 kPa, low NS
LN, 90 kPa, high vs. FN, 90 kPa, low NS
LN, 90 kPa, high vs. LN, 90 kPa, low NS
FN, 13 kPa, low vs. LN, 13 kPa, low NS
FN, 13 kPa, low vs. FN, 90 kPa, low NS
FN, 13 kPa, low vs. LN, 90 kPa, low NS
LN, 13 kPa, low vs. FN, 90 kPa, low NS
LN, 13 kPa, low vs. LN, 90 kPa, low NS
FN, 90 kPa, low vs. LN, 90 kPa, low NS

93

Supplementary Table S5. Statistical analysis for average time to peak systole in
µMyocardia.
All data was not normally distributed, as determined by the Lilliefors test. The p-value for the
Kruskal-Wallis test was 0.0121, the chi-square value was 17.97, and there were 7 degrees of
freedom. Multiple comparisons were performed using the Bonferroni method, with p values for
statistical differences indicated in the table below. The statistical power is 62.89%.

Comparison p-value
FN, 13 kPa, high vs. LN, 13 kPa, high NS
FN, 13 kPa, high vs. FN, 90 kPa, high NS
FN, 13 kPa, high vs. LN, 90 kPa, high NS
FN, 13 kPa, high vs. FN, 13 kPa, low NS
FN, 13 kPa, high vs. LN, 13 kPa, low NS
FN, 13 kPa, high vs. FN, 90 kPa, low NS
FN, 13 kPa, high vs. LN, 90 kPa, low NS
LN, 13 kPa, high vs. FN, 90 kPa, high NS
LN, 13 kPa, high vs. LN, 90 kPa, high NS
LN, 13 kPa, high vs. FN, 13 kPa, low NS
LN, 13 kPa, high vs. LN, 13 kPa, low NS
LN, 13 kPa, high vs. FN, 90 kPa, low NS
LN, 13 kPa, high vs. LN, 90 kPa, low NS
FN, 90 kPa, high vs. LN, 90 kPa, high NS
FN, 90 kPa, high vs. FN, 13 kPa, low NS
FN, 90 kPa, high vs. LN, 13 kPa, low NS
FN, 90 kPa, high vs. FN, 90 kPa, low NS
FN, 90 kPa, high vs. LN, 90 kPa, low NS
LN, 90 kPa, high vs. FN, 13 kPa, low NS
LN, 90 kPa, high vs. LN, 13 kPa, low NS
LN, 90 kPa, high vs. FN, 90 kPa, low NS
LN, 90 kPa, high vs. LN, 90 kPa, low NS
FN, 13 kPa, low vs. LN, 13 kPa, low NS
FN, 13 kPa, low vs. FN, 90 kPa, low NS
FN, 13 kPa, low vs. LN, 90 kPa, low NS
LN, 13 kPa, low vs. FN, 90 kPa, low NS
LN, 13 kPa, low vs. LN, 90 kPa, low NS
FN, 90 kPa, low vs. LN, 90 kPa, low NS

94

Supplementary Table S6. Statistical analysis for peak systolic longitudinal displacement in
µMyocardia.
All data was not normally distributed, as determined by the Lilliefors test. The p-value for the
Kruskal-Wallis test was 6.99e-30, the chi-square value was 153.64, and there were 7 degrees of
freedom. Multiple comparisons were performed using the Bonferroni method, with p values for
statistical differences indicated in the table below. The statistical power is 100%.

Comparison p-value
FN, 13 kPa, high vs. LN, 13 kPa, high NS
FN, 13 kPa, high vs. FN, 90 kPa, high p<0.05
FN, 13 kPa, high vs. LN, 90 kPa, high p<0.05
FN, 13 kPa, high vs. FN, 13 kPa, low NS
FN, 13 kPa, high vs. LN, 13 kPa, low NS
FN, 13 kPa, high vs. FN, 90 kPa, low p<0.05
FN, 13 kPa, high vs. LN, 90 kPa, low p<0.05
LN, 13 kPa, high vs. FN, 90 kPa, high p<0.05
LN, 13 kPa, high vs. LN, 90 kPa, high p<0.05
LN, 13 kPa, high vs. FN, 13 kPa, low NS
LN, 13 kPa, high vs. LN, 13 kPa, low NS
LN, 13 kPa, high vs. FN, 90 kPa, low p<0.05
LN, 13 kPa, high vs. LN, 90 kPa, low p<0.05
FN, 90 kPa, high vs. LN, 90 kPa, high NS
FN, 90 kPa, high vs. FN, 13 kPa, low p<0.05
FN, 90 kPa, high vs. LN, 13 kPa, low p<0.05
FN, 90 kPa, high vs. FN, 90 kPa, low NS
FN, 90 kPa, high vs. LN, 90 kPa, low NS
LN, 90 kPa, high vs. FN, 13 kPa, low p<0.05
LN, 90 kPa, high vs. LN, 13 kPa, low p<0.05
LN, 90 kPa, high vs. FN, 90 kPa, low NS
LN, 90 kPa, high vs. LN, 90 kPa, low NS
FN, 13 kPa, low vs. LN, 13 kPa, low NS
FN, 13 kPa, low vs. FN, 90 kPa, low p<0.05
FN, 13 kPa, low vs. LN, 90 kPa, low p<0.05
LN, 13 kPa, low vs. FN, 90 kPa, low p<0.05
LN, 13 kPa, low vs. LN, 90 kPa, low p<0.05
FN, 90 kPa, low vs. LN, 90 kPa, low NS

95

Supplementary Table S7. Statistical analysis for peak systolic longitudinal cross-sectional
force (F
x
) in µMyocardia.
All data was not normally distributed, as determined by the Lilliefors test. The p-value for the
Kruskal-Wallis test was 1.18e-12, the chi-square value was 70.48, and there were 7 degrees of
freedom. Multiple comparisons were performed using the Bonferroni method, with p values for
statistical differences indicated in the table below. The statistical significance is 99.87%.

Comparison p-value
FN, 13 kPa, high vs. LN, 13 kPa, high NS
FN, 13 kPa, high vs. FN, 90 kPa, high NS
FN, 13 kPa, high vs. LN, 90 kPa, high p<0.05
FN, 13 kPa, high vs. FN, 13 kPa, low NS
FN, 13 kPa, high vs. LN, 13 kPa, low p<0.05
FN, 13 kPa, high vs. FN, 90 kPa, low p<0.05
FN, 13 kPa, high vs. LN, 90 kPa, low p<0.05
LN, 13 kPa, high vs. FN, 90 kPa, high NS
LN, 13 kPa, high vs. LN, 90 kPa, high p<0.05
LN, 13 kPa, high vs. FN, 13 kPa, low NS
LN, 13 kPa, high vs. LN, 13 kPa, low p<0.05
LN, 13 kPa, high vs. FN, 90 kPa, low p<0.05
LN, 13 kPa, high vs. LN, 90 kPa, low p<0.05
FN, 90 kPa, high vs. LN, 90 kPa, high NS
FN, 90 kPa, high vs. FN, 13 kPa, low NS
FN, 90 kPa, high vs. LN, 13 kPa, low NS
FN, 90 kPa, high vs. FN, 90 kPa, low NS
FN, 90 kPa, high vs. LN, 90 kPa, low p<0.05
LN, 90 kPa, high vs. FN, 13 kPa, low NS
LN, 90 kPa, high vs. LN, 13 kPa, low NS
LN, 90 kPa, high vs. FN, 90 kPa, low NS
LN, 90 kPa, high vs. LN, 90 kPa, low NS
FN, 13 kPa, low vs. LN, 13 kPa, low p<0.05
FN, 13 kPa, low vs. FN, 90 kPa, low NS
FN, 13 kPa, low vs. LN, 90 kPa, low p<0.05
LN, 13 kPa, low vs. FN, 90 kPa, low NS
LN, 13 kPa, low vs. LN, 90 kPa, low NS
FN, 90 kPa, low vs. LN, 90 kPa, low NS

96

Supplementary Table S8. Statistical analysis for peak systolic work (W) in µMyocardia.
All data was not normally distributed, as determined by the Lilliefors test. The p-value for the
Kruskal-Wallis test was 9.144e-30, the chi-square value was 153.08, and there were 7 degrees of
freedom. Multiple comparisons were performed using the Bonferroni method, with p values for
statistical differences indicated in the table below. The statistical power is 100%.

Comparison p-value
FN, 13 kPa, high vs. LN, 13 kPa, high NS
FN, 13 kPa, high vs. FN, 90 kPa, high p<0.05
FN, 13 kPa, high vs. LN, 90 kPa, high p<0.05
FN, 13 kPa, high vs. FN, 13 kPa, low NS
FN, 13 kPa, high vs. LN, 13 kPa, low NS
FN, 13 kPa, high vs. FN, 90 kPa, low p<0.05
FN, 13 kPa, high vs. LN, 90 kPa, low p<0.05
LN, 13 kPa, high vs. FN, 90 kPa, high p<0.05
LN, 13 kPa, high vs. LN, 90 kPa, high p<0.05
LN, 13 kPa, high vs. FN, 13 kPa, low NS
LN, 13 kPa, high vs. LN, 13 kPa, low NS
LN, 13 kPa, high vs. FN, 90 kPa, low p<0.05
LN, 13 kPa, high vs. LN, 90 kPa, low p<0.05
FN, 90 kPa, high vs. LN, 90 kPa, high NS
FN, 90 kPa, high vs. FN, 13 kPa, low p<0.05
FN, 90 kPa, high vs. LN, 13 kPa, low p<0.05
FN, 90 kPa, high vs. FN, 90 kPa, low NS
FN, 90 kPa, high vs. LN, 90 kPa, low NS
LN, 90 kPa, high vs. FN, 13 kPa, low p<0.05
LN, 90 kPa, high vs. LN, 13 kPa, low p<0.05
LN, 90 kPa, high vs. FN, 90 kPa, low NS
LN, 90 kPa, high vs. LN, 90 kPa, low NS
FN, 13 kPa, low vs. LN, 13 kPa, low NS
FN, 13 kPa, low vs. FN, 90 kPa, low p<0.05
FN, 13 kPa, low vs. LN, 90 kPa, low p<0.05
LN, 13 kPa, low vs. FN, 90 kPa, low p<0.05
LN, 13 kPa, low vs. LN, 90 kPa, low p<0.05
FN, 90 kPa, low vs. LN, 90 kPa, low NS

97

Supplementary Table S9. Statistical analysis for peak systolic lateral displacement in
µMyocardia.
All data was not normally distributed, as determined by the Lilliefors test. The p-value for the
Kruskal-Wallis test was 1.834e-30, the chi-square value was 156.4, and there were 7 degrees of
freedom. Multiple comparisons were performed using the Bonferroni method, with p values for
statistical differences indicated in the table below. The statistical power is 99.99%.
Comparison p-value
FN, 13 kPa, high vs. LN, 13 kPa, high NS
FN, 13 kPa, high vs. FN, 90 kPa, high p<0.05
FN, 13 kPa, high vs. LN, 90 kPa, high p<0.05
FN, 13 kPa, high vs. FN, 13 kPa, low NS
FN, 13 kPa, high vs. LN, 13 kPa, low NS
FN, 13 kPa, high vs. FN, 90 kPa, low p<0.05
FN, 13 kPa, high vs. LN, 90 kPa, low p<0.05
LN, 13 kPa, high vs. FN, 90 kPa, high p<0.05
LN, 13 kPa, high vs. LN, 90 kPa, high p<0.05
LN, 13 kPa, high vs. FN, 13 kPa, low NS
LN, 13 kPa, high vs. LN, 13 kPa, low NS
LN, 13 kPa, high vs. FN, 90 kPa, low p<0.05
LN, 13 kPa, high vs. LN, 90 kPa, low p<0.05
FN, 90 kPa, high vs. LN, 90 kPa, high NS
FN, 90 kPa, high vs. FN, 13 kPa, low p<0.05
FN, 90 kPa, high vs. LN, 13 kPa, low p<0.05
FN, 90 kPa, high vs. FN, 90 kPa, low NS
FN, 90 kPa, high vs. LN, 90 kPa, low NS
LN, 90 kPa, high vs. FN, 13 kPa, low p<0.05
LN, 90 kPa, high vs. LN, 13 kPa, low p<0.05
LN, 90 kPa, high vs. FN, 90 kPa, low NS
LN, 90 kPa, high vs. LN, 90 kPa, low NS
FN, 13 kPa, low vs. LN, 13 kPa, low NS
FN, 13 kPa, low vs. FN, 90 kPa, low p<0.05
FN, 13 kPa, low vs. LN, 90 kPa, low p<0.05
LN, 13 kPa, low vs. FN, 90 kPa, low p<0.05
LN, 13 kPa, low vs. LN, 90 kPa, low p<0.05
FN, 90 kPa, low vs. LN, 90 kPa, low NS

98

Supplementary Table S10. Statistical analysis for peak systolic lateral cross-sectional force
(F
y
) in µMyocardia.
All data was not normally distributed, as determined by the Lilliefors test. The p-value for the
Kruskal-Wallis test was 0.0008, the chi-square value was 24.91, and there were 7 degrees of
freedom. Multiple comparisons were performed using the Bonferroni method, with p values for
statistical differences indicated in the table below. The statistical power is 80.40%.

Comparison p-value
FN, 13 kPa, high vs. LN, 13 kPa, high NS
FN, 13 kPa, high vs. FN, 90 kPa, high NS
FN, 13 kPa, high vs. LN, 90 kPa, high NS
FN, 13 kPa, high vs. FN, 13 kPa, low NS
FN, 13 kPa, high vs. LN, 13 kPa, low NS
FN, 13 kPa, high vs. FN, 90 kPa, low NS
FN, 13 kPa, high vs. LN, 90 kPa, low NS
LN, 13 kPa, high vs. FN, 90 kPa, high NS
LN, 13 kPa, high vs. LN, 90 kPa, high NS
LN, 13 kPa, high vs. FN, 13 kPa, low NS
LN, 13 kPa, high vs. LN, 13 kPa, low NS
LN, 13 kPa, high vs. FN, 90 kPa, low NS
LN, 13 kPa, high vs. LN, 90 kPa, low NS
FN, 90 kPa, high vs. LN, 90 kPa, high NS
FN, 90 kPa, high vs. FN, 13 kPa, low NS
FN, 90 kPa, high vs. LN, 13 kPa, low p<0.05
FN, 90 kPa, high vs. FN, 90 kPa, low NS
FN, 90 kPa, high vs. LN, 90 kPa, low p<0.05
LN, 90 kPa, high vs. FN, 13 kPa, low NS
LN, 90 kPa, high vs. LN, 13 kPa, low NS
LN, 90 kPa, high vs. FN, 90 kPa, low NS
LN, 90 kPa, high vs. LN, 90 kPa, low NS
FN, 13 kPa, low vs. LN, 13 kPa, low NS
FN, 13 kPa, low vs. FN, 90 kPa, low NS
FN, 13 kPa, low vs. LN, 90 kPa, low NS
LN, 13 kPa, low vs. FN, 90 kPa, low NS
LN, 13 kPa, low vs. LN, 90 kPa, low NS
FN, 90 kPa, low vs. LN, 90 kPa, low NS

99

Supplemental 2: Aim 3

Supplementary Figure S2. Control of Sectioned Muscle Bundles. Control sections
immunostained with only the secondary antibodies for 8x2x2 (A-C) and 13x2x2 (D-F) mm
bundles. Merged images of DAPI, Alexa Fluor 546 goat anti-mouse secondary antibody, and
Alexa Fluor 488 goat anti-rabbit secondary antibody (A,D) and images of Alexa Fluor 546 goat
anti-mouse secondary antibody (B,E) and Alexa Fluor 488 goat anti-rabbit secondary antibody
(C,F) are shown. Scale bars, 150 µm; blue, nuclei; green, dystrophin; red, sarcomeric α-actinin.

100

Supplementary Figure S3. Control of Cleared Muscle Bundles. Control cleared bundles
immunostained with only the secondary antibodies for 8x2x2 mm bundles. Merged images of
DAPI, Alexa Fluor 546 goat anti-mouse secondary antibody, and Alexa Fluor 488 goat anti-rabbit
secondary antibody (A) and images of Alexa Fluor 546 goat anti-mouse secondary antibody
(B,E) and Alexa Fluor 488 goat anti-rabbit secondary antibody (C) are shown. Scale bars, 150
µm; blue, nuclei; green, dystrophin; red, sarcomeric α-actinin.

101

Supplementary Table S11. Statistical analysis for percentage bundle survival.
Cox Proportional Hazards Regression was used to compare the survival curves. The coefficient
b, standard error, p-value, and hazard ratio (e
b
) are indicated in the table below.

Coefficient (b) standard error p e
b

13x2x2 vs 8x2x2 -7.07E-17 0.4714 1 1
13x2x2 vs 8x1x2 0.3204 0.2383 0.1788 1.377679
13x2x2 vs 8x2x1 0.2157 0.1587 0.174 1.24073
8x2x2 vs 8x1x2 0.6408 0.4765 0.1788 1.897999
8x2x2 vs 8x2x1 0.3236 0.238 0.174 1.382094
8x1x2 vs 8x2x1 0.0996 0.4724 0.833 1.104729

Supplementary Table S12. Statistical analysis for change in bundle width.
All data was not normally distributed, as determined by the Lilliefors test. The p-value for the
Kruskal-Wallis test, the chi-square value, and the 7 degrees of freedom are listed below. If
p<0.05 for the Kruskal-Wallis test, multiple comparisons were performed using the Bonferroni
method, with p values for statistical differences indicated in the table below.
Day 1 Day 2 Day 4 Day 6 Day 8 Day 10 Day 12
(p=
1.218e-7,
χ
2
= 35,
Degrees
of
freedom =
35)
(p=
0.0039,
χ
2
= 13.38,
Degrees
of
freedom =
35)
(p=
0.0007,
χ
2
= 17.1,
Degrees
of
freedom
= 32)
(p=
0.0002,
χ
2
= 20.0
7,
Degrees
of
freedom
= 31)
(p=
0.0004,
χ
2
= 18.09,
Degrees of
freedom =
27)
(p=
0.0011,
χ
2
= 16.08,
Degrees of
freedom =
26)
(p= 0.0047,
χ
2
= 12.99,
Degrees of
freedom =
25)
13x2x
2 vs
8x2x2
NS NS NS NS NS NS NS
13x2x
2 vs
8x1x2
p<0.05 NS p<0.05 p<0.05 p<0.05 p<0.05 p<0.05
13x2x
2 vs
8x2x1
NS NS NS NS NS NS NS
8x2x2
vs
8x1x2
p<0.05 p<0.05 p<0.05 p<0.05 p<0.05 NS NS
8x2x2
vs
8x2x1
NS NS NS NS NS NS NS
8x1x2
vs
8x2x1
p<0.05 p<0.05 p<0.05 NS NS NS NS

102

Day 14 Day 16 Day 18 Day 20 Day 22 Day 24
(p=
0.0033,
χ
2
= 13.74,
Degrees of
freedom =
25)
(p=
0.0183,
χ
2
= 10.03,
Degrees of
freedom =
25)
(p=
0.0282,
χ
2
= 9.09,
Degrees
of
freedom
= 24)
(p=
0.0665,
χ
2
= 7.18,
Degrees of
freedom =
24)
(p=
0.0188,
χ
2
= 9.97,
Degrees
of
freedom
= 24)
(p=
0.03,
χ
2
= 8.95,
Degrees
of
freedom
= 24)
13x2x
2 vs
8x2x2
NS NS NS Not
applicable
NS NS
13x2x
2 vs
8x1x2
p<0.05 NS NS Not
applicable
NS NS
13x2x
2 vs
8x2x1
p<0.05 NS NS Not
applicable
NS NS
8x2x2
vs
8x1x2
NS NS NS Not
applicable
NS NS
8x2x2
vs
8x2x1
NS NS NS Not
applicable
NS NS
8x1x2
vs
8x2x1
NS NS NS Not
applicable
NS NS

Abstract (if available)

Abstract Striated muscle diseases, including cardiovascular diseases and muscular dystrophies, are a significant burden on the United States. Our ability to develop therapies for these diseases is hampered by a reliance on models systems that cannot accurately and efficiently replicate human striated muscle tissue and/or quantify its phenotype. Also, striated muscle diseases can be either acquired or genetic, which present unique challenges when modeling these different forms. To overcome these challenges, tissue engineering has recently been utilized to develop more sophisticated models of striated muscle, but many of these models still lack physiological relevance, scalability, reproducibility, and functional outputs. Here, our goal was to engineer scalable striated muscle tissue platforms that can be used to determine the effects of disease-relevant microenvironmental factors and/or genetic mutations on muscle phenotype. First, we engineered a 2D platform to quantify the effects of cellular and extracellular remodeling on the contractility of 2D cardiac tissues. We will also engineer 3D skeletal muscle tissues using 3D printing for rapid prototyping and tissue clearing for enhanced visualization. We established our approach using C2C12 mouse myoblasts by determining the effect of mold dimensions on the percentage survival and change in bundle width of 3D bundles. We also demonstrated that bundles respond to electrical stimulation by contracting. Last, we used tissue clearing to visualize the bundles in multiple planes. The platforms developed here can be used to create “organs-on-chips” mimicking the structure and function of the heart and skeletal muscle. These chips could be used to create high-throughput quantification of organ function and organ interactions to drugs.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Engineering 2D & 3D microphysiological systems for interrogating skeletal muscle tissues

PDF

New approaches for precisely engineering heterotypic muscle tissues by naturally and synthetically controlling cell fate

PDF

Engineering a post-infarct human myocardium on a chip to reveal oxygen-dependent crosstalk in cardiac fibroblasts and myocytes

PDF

Spatially controlled tissue differentiation using the synthetic receptor SynNotch

PDF

Extracellular matrix regulation of mitochondrial function in engineered cardiac myocytes

PDF

Understanding anti-angiogenic signaling and treatment for cancer through mechanistic modeling

PDF

Micelle nanoparticles for the detection and treatment of atherosclerosis

PDF

Application of small molecules as self-reprogramming agents for the reconstruction of musculoskeletal tissues within 3-D tissue-specific scaffolds

PDF

The novel material: organosilanes crosslinked gelatin its characteristics and potentials for tissue engineering applications

PDF

Physiologically-based modeling to predict monoclonal antibody pharmacokinetics in humans from physiochemical properties

PDF

Regulation of linear and angular impulse generation: implications for athletic performance

PDF

Metabolic profiling of single hematopoietic stem cells for developing novel ex vivo culture strategies

PDF

Knowledge‐driven decision support for assessing radiation therapy dose constraints

PDF

Development and implementation of a modular muscle-computer interface for personalized motor rehabilitation after stroke

PDF

Pharmacokinetic/pharmacodynamic modeling for genetically polymorphic populations

PDF

Face recognition and 3D face modeling from images in the wild

PDF

A parallel computation framework for EONS synaptic modeling platform for parameter optimization and drug discovery

PDF

Modeling human regulation of momentum while interacting with the environment

PDF

Systems pharmacology for acute myeloid leukemia and feedback control pharmacodynamic models

PDF

Learning controllable data generation for scalable model training

Asset Metadata

Creator Ariyasinghe, Nethika Ruvini (author)

Core Title Engineering scalable two- and three-dimensional striated muscle microtissues for human disease modeling

Contributor Electronically uploaded by the author (provenance)

School Andrew and Erna Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Biomedical Engineering

Publication Date 01/11/2020

Defense Date 05/07/2019

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag 3D printing,cardiac muscle,extracellular matrix,myocardial infarction,OAI-PMH Harvest,organs-on-chips,skeletal muscle,tissue clearing,traction force microscopy

Format application/pdf (imt)

Language English

Advisor McCain, Megan Laura (committee chair), D'Argenio, David (committee member), McNitt-Gray, Jill (committee member)

Creator Email nariyasi@usc.edu,nariyasinghe@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-182187

Unique identifier UC11660084

Identifier etd-Ariyasingh-7540.pdf (filename),usctheses-c89-182187 (legacy record id)

Legacy Identifier etd-Ariyasingh-7540.pdf

Dmrecord 182187

Document Type Dissertation

Format application/pdf (imt)

Rights Ariyasinghe, Nethika Ruvini

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA