University of Southern California Dissertations and Theses

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

(USC Thesis Other)

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

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Copyright 2024 Natalie N. Khalil
ENGINEERING A POST-INFARCT HUMAN MYOCARDIUM ON A CHIP TO REVEAL
OXYGEN-DEPENDENT CROSSTALK IN CARDIAC FIBROBLASTS AND MYOCYTES
by
Natalie N. Khalil
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(BIOMEDICAL ENGINEERING)
May 2024

ii
Acknowledgements
I thank my advisor, Dr. Megan McCain, for providing me with the guidance and mentorship
necessary to become an independent researcher. I admire her ability to lead a growing laboratory
and department. It has been remarkable to be a part of the lab as it evolved from its initial focus
on skeletal and cardiac muscle, to new prospects within neural engineering, synthetic biology,
and women’s health. I am excited to keep up-to-date with her progressive vision for the laboratory.
In addition, I extend my appreciation to the following individuals:
1. My Dissertation Committee, Dr. Eunji Chung, Dr. Ellen Lien, Dr. Shannon Mumenthaler,
and Dr. Keyue Shen. I am grateful for their time, feedback, and guidance.
2. The Laboratory of Dr. Sarah J. Parker at Cedars-Sinai Medical Center. I extend my
appreciation to Dr. Parker for her mentorship, and to her laboratory, including Divya Gupta,
Liam McCarthy, and Sean Escopete, for their integral contributions to this work.
3. The Laboratory for Living Systems Engineering. I am thankful to have learned from lab
alumni Davi Lyra-Leite, Nethika Ariyasinghe, Nathan Cho, Andrew Petersen, Joycelyn Yip,
and Jeffrey Santoso, as well as postdoctoral researcher Megan Rexius-Hall. Thanks also
to current lab members Alisa Peshina, James Eichenbaum, Mher Garibyan, Nina Maxey,
Riya Verma and Steph Do for their friendship as we navigated the program together.
4. My parents Imane and Nouhad, older brothers Shadi and Nader, and partner Matteo. My
parents immigrated to America to escape a war with little working knowledge of the English
language. Their stories have inspired my work ethic and appreciation for life. Daily phone
calls with my mom during my commute to and from lab imparted in me her wisdom and
stoicism, and taught me the importance of kindness, humility, and self-reflection. Shadi
serves as an academic role model for me in his continuous pursuit of new knowledge,
while Nader’s ambition and entrepreneurial spirit is a constant reminder that I can achieve
my dreams. Lastly, Matteo has inspired me to remain rational and grounded through
adversity and has provided unwavering love and support. I understand how lucky I am to
have such an encouraging family.

iii
Table of Contents
Acknowledgements.....................................................................................................................ii
List of Figures ............................................................................................................................vi
Abstract.......................................................................................................................................x
Chapter 1 Introduction: Engineering the Cellular Microenvironment of Post-Infarct Myocardium…..
on a Chip ................................................................................................................................... 1
1.1. The Clinical Perspective....................................................................................... 2
1.1.1. Clinical Diagnoses ......................................................................................... 2
1.1.2. Current Treatments ....................................................................................... 3
1.2. Healing After Myocardial Infarction....................................................................... 5
1.2.1. Tissue-Level Response: Ventricular Remodeling........................................... 5
1.2.2. Cellular and Molecular Response: Phases of Wound Healing ....................... 7
1.2.3. The Pathological Fibrotic Response .............................................................. 9
1.3. In Vitro Hypoxia Models ......................................................................................11
1.3.1. Conventional Hypoxia Models ......................................................................12
1.3.2. Engineered Models: Spatially Controlled Oxygen .........................................14
1.3.3. Engineered Models: Temporally Controlled Oxygen .....................................16
1.4. In Vitro Stiffness Models .....................................................................................18
1.4.1. Two-Dimensional Models with Uniform Stiffness ..........................................19
1.4.2. Two-Dimensional Models with Spatiotemporal Control over Stiffness...........22
1.4.3. Three-Dimensional Models with Uniform Stiffness........................................23
1.4.4. Three-Dimensional Models with Spatiotemporal Control over Stiffness ........25
1.5. In Vitro Models of Pathological Strain..................................................................27
1.6. Objective and Aims .............................................................................................31
Chapter 2 Human Cardiac Fibroblast Phenotypes Induced by Acute Hypoxia ..........................32
2.1. Introduction .........................................................................................................33
2.2. Materials and Methods........................................................................................34
2.2.1. Fabrication of PDMS Substrates...................................................................34
2.2.2. Cell Culture...................................................................................................35
2.2.3. Immunostaining ............................................................................................36
2.2.4. Microscopy and Image Analysis ...................................................................36
2.2.5. RNA and Protein Isolation ............................................................................37
2.2.6. Quantitative RT-PCR....................................................................................37
2.2.7. Proteomic Analysis .......................................................................................38
2.3. Results................................................................................................................39
2.3.1. Hypoxia Drives Nuclear Translocation of HIF-1α within Hours......................39
2.3.2. Hypoxia Inhibits Cardiac Fibroblast Proliferation...........................................40
2.3.3. Hypoxia Drives Expression of the HIF Target Gene VEGFA.........................41

iv
2.3.4. Hypoxia Downregulates Collagen Synthesis.................................................42
2.3.5. Hypoxia Globally Represses Protein Synthesis in Cardiac Fibroblasts .........43
2.4. Discussion ..........................................................................................................47
Chapter 3 Hypoxic-Normoxic Crosstalk under Microfluidic Control Activates Unique…..
Inflammatory Pathways in Human Cardiac Fibroblasts .............................................................51
3.1. Introduction .........................................................................................................51
3.2. Materials and Methods........................................................................................53
3.2.1. Microfluidic Device Fabrication .....................................................................53
3.2.2. Cell Culture Chamber Fabrication.................................................................54
3.2.3. Oxygen Validation ........................................................................................57
3.2.4. Cell Culture...................................................................................................57
3.2.5. Microfluidic Oxygen Modulation ....................................................................58
3.2.6. Immunostaining ............................................................................................58
3.2.7. Microscopy and Image Analysis ...................................................................59
3.2.8. RNA and Cell Media Isolation.......................................................................59
3.2.9. RNA-Sequencing and Data Analysis ............................................................60
3.2.10. MSD Assay.................................................................................................61
3.3. Results................................................................................................................61
3.3.1. Microfluidic Devices Enable Co-Culture of Hypoxic and Normoxic Cells .......61
3.3.2. HIF Activation is Restricted to Hypoxic Cardiac Fibroblasts Regardless of……
Co-Culture .........................................................................................................................63
3.3.3. Co-Culture with Normoxic Fibroblasts Activates Pro-Inflammatory..….
Phenotypes in Hypoxic Fibroblasts ....................................................................................68
3.4. Discussion ..........................................................................................................71
Chapter 4 A Post-Infarct Human Myocardium on a Chip Reveals Hypoxic Cardiac …….
Fibroblasts Drive Cytokine Signaling in Normoxic Cardiac Myocytes ........................................74
4.1. Introduction .........................................................................................................74
4.2. Materials and Methods........................................................................................76
4.2.1. Microfluidic Device Fabrication .....................................................................76
4.2.2. Cell Culture Chamber and Co-Seeding Chamber Fabrication.......................77
4.2.3. Dual-Seeding................................................................................................77
4.2.4. Human Cardiac Fibroblast Cell Culture.........................................................77
4.2.5. Human iPSC-derived Cardiac Myocyte Differentiation and Cell Culture........78
4.2.6. Oxygen Modulation.......................................................................................79
4.2.7. Immunostaining and Image Analysis ............................................................79
4.2.8. RNA and Protein Isolation ............................................................................79
4.2.9. RNA-Sequencing and Data Analysis ............................................................80
4.2.10. MSD Assay.................................................................................................80
4.3. Results................................................................................................................81
4.3.1. A 3D-Printed and Replica-Molded Chamber Enables Dual Seeding of….…
Human Cardiac Fibroblasts and Human iPSC-Derived Cardiac Myocytes .........................81

v
4.3.2. Hypoxic Cardiac Fibroblasts Impair Neighboring Normoxic Cardiac……
Myocyte Cell Counts and Morphology................................................................................82
4.3.3. RNA-Sequencing Demonstrates Primary Clustering by Cell Type ................85
4.3.4. Normoxic Cardiac Myocytes Increase Cytokine Signaling After………………
Co-Culture with Hypoxic Cardiac Fibroblasts .....................................................................86
4.3.5. Hypoxic cardiac fibroblasts are a more prolific source of cytokines………….
than normoxic cardiac myocytes ........................................................................................88
4.4. Conclusions ........................................................................................................91
Chapter 5 Outlook.....................................................................................................................93
5.1. Fibroblasts as Therapeutic Targets .....................................................................94
5.2. Towards a 4-D Human Myocardium on a Chip....................................................96
5.3. Final Conclusions................................................................................................98
Resources...............................................................................................................................100
References .............................................................................................................................101

vi
List of Figures
Figure 1-1. The post-infarct microenvironment is characterized by dynamic physical, mechanical,
and biochemical cues that orchestrate the three phases of healing. Regional and temporal
changes of oxygen concentration, stiffness, inflammatory cytokines, cell and matrix composition
characterize the myocardium as injured tissue is replaced with a scar.
Created with BioRender.com…………………………………………………………………………..11
Figure 1-2. Engineered in vitro hypoxia models that modulate oxygen in (A) space and (B) time.
(A) Infarct spheroids generate intrinsic spatial oxygen gradients to model the infarct, border, and
remote zones after myocardial infarction. Infarct spheroids exhibit higher levels of myofibroblast
marker expression, exhibit asynchronous contraction of myocytes, and have lower contraction
amplitudes than control spheroids. Reprinted by permission from Springer Nature: Nat. Biomed.
Eng., (Richards et al., 2020). (B) Muscular thin films with integrated strain gauges can measure
changes in contractility in cardiac myocytes pre-treated with endothelial extracellular vesicles in
a simulated ischemia-reperfusion injury. Endothelial extracellular vesicles are cardioprotective
and provide cardiac tissues with enhanced contractility during both ischemia and reperfusion.
From (Yadid et al., 2020). Reprinted with permission from AAAS…………………………………18
Figure 1-3. Engineered in vitro fibrosis models that modulate stiffness in (A) time and (B) space.
(A) A hyaluronic acid hydrogel is polymerized in situ upon exposure to UV light to mimic scar
maturation after myocardial infarction. Stiffening from 8 to 30 kPa results in collagen turnover.
Republished with permission of the American Society for Cell Biology, from (Herum et al., 2017a);
permission conveyed through Copyright Clearance Center, Inc. (B) A heteropolar biowire
integrates normal (7.6 kPa) and fibrotic (61.1 kPa) cardiac tissue. The white dashed line indicates
the interface between normal and fibrotic tissue. Normal regions exhibit increased α-actinin, a
component of sarcomeres in cardiac myocytes, while fibrotic regions have increased collagen
and α-smooth muscle actin, markers of myofibroblasts. Images reproduced with permission of the
American Chemical Society, from https://pubs.acs.org/doi/full/10.1021/acscentsci.9b00052
(Wang et al., 2019). Further permissions related to the material excerpted should be directed to
the American Chemical
Society……………………………………………………………………………………………………27
Figure 1-4. Engineered in vitro models to mimic pathological strain post-infarction. (A) Muscular
thin films on a stretchable silicone membrane provide real-time measurements of contractile
stress during stretch. Images reproduced from (McCain et al., 2013). (B) A microfluidic device to
evaluate the combined effects of hypoxia (yellow channel) and strain (red channel) on cardiac
fibroblasts. Created with BioRender.com. This work demonstrates that the combined effects of
hypoxia (1% O2) and reduced contractility (2% strain) to mimic injured myocardium are required
for cardiac fibroblasts to upregulate TGF-β and IL-1β. Image reproduced from (Ugolini et al.,
2017)(https://creativecommons.org/licenses/by/4.0/)..................................................................30
Figure 2-1. HIF-1 and HIF-2 localization in human cardiac fibroblasts treated with hypoxia or TGFβ1. Immunofluorescence of (A) HIF-1 (red) and DAPI (blue) and quantification of HIF-1
nuclear/cytosolic intensity. (B) HIF-2 (red) and DAPI (blue) and quantification of HIF-2
nuclear/cytosolic intensity. Brightness and contrast are adjusted on all images. Scale bars = 100
µm. NT = no treatment. Each data point is the average of 5 images per coverslip across n=4
coverslips from 3 independent trials. ****p<0.0001 according to ordinary one-way ANOVA with
Dunnett’s multiple comparisons with NT as the control……………………………………………..40

vii
Figure 2-2. Differentiation and proliferation of human cardiac fibroblasts treated with hypoxia or
TGF-β1. (A) α-SMA (red), actin (green), DAPI (blue) and quantification of α-SMA+ cells. (B) Ki67 (green) and DAPI (blue) and quantification of Ki67+ nuclei. Brightness and contrast are
adjusted on all images. Scale bars = 100 µm. NT = no treatment. Each data point is the average
of 5 images per coverslip across n=4 coverslips from 3 independent trials. *p<0.05, **p<0.01,
according to ordinary one-way ANOVA with Dunnett’s multiple comparisons with NT as the
control……………………………………………………………………………………...……………..41
Figure 2-3. Expression of hypoxic and fibrotic target genes in human cardiac fibroblasts treated
with hypoxia or TGF-β1. Quantitative RT-PCR of genes encoding (A) vimentin, (B) vascular
endothelial growth factor, (C) alpha-smooth muscle actin, (D) collagen type I, and (E) transforming
growth factor β1 after hypoxia (0% O2) or TGF-β1 treatment. NT = no treatment. Each data point
reflects two pooled coverslips per trial from n=4 independent trials. *p<0.05, ***p<0.001,
****p<0.0001 according to ordinary one-way ANOVA with Dunnett’s multiple comparisons with NT
as the control……………………………………………………………………………………………..42
Figure 2-4. Proteomic analysis of human cardiac fibroblasts treated with hypoxia or TGF-β1. (A)
Venn diagram comparing significant proteins after hypoxia or TGF-β1 treatment. Normalized
intensities for (B) TGFB1, (C) CO1A1, (D) CO1A2, (E) YAP1, and (D) LOX. NT = no treatment.
Each data point reflects one of n=5 or 6 coverslips from 4 independent trials. *p<0.05, ***p<0.001,
****p<0.0001 according to ordinary one-way ANOVA with Dunnett’s multiple comparisons with NT
as the control……………………………………………………………………………………………..43
Figure 2-5. Pathway analysis of human cardiac fibroblast proteomes after treatment with 24h
hypoxia or TGF-β1. Top 20 pathways, sorted by p-value, in cells treated with (A) hypoxia or (B)
TGF-β1, after applying p-value and absolute z-score cutoffs of 0.05 and 1.5, respectively. (C) Dot
plot reflecting shared pathways that pass p-value cutoffs of 0.05 in both conditions, where size
represents significance and color denotes z-score………………………………………………….45
Figure 2-6. Expression of proteins with highest significance or fold change in human cardiac
fibroblasts treated with hypoxia or TGF-β1. Normalized intensites of (A) Angiopoietin-like 4, (B)
ferritin heavy chain 1, (C) N-myc downstream regulated gene 1, (D) connective tissue growth
factor, (E) palladin, (F) acetyl co-A binding protein, (G) pyruvate kinase L/R, (H) LanC like
glutathione S transferase 2, and (I) prion protein. NT = no treatment. Each data point is one of
n=5 or 6 coverslips from 4 independent trials. *p<0.05, ***p<0.001, ****P<0.0001 according to
ordinary one-way ANOVA with Dunnett’s multiple comparisons with NT as the control………..47
Figure 3-1. (A) Digital light processing 3D-printing enables rapid prototyping of microfluidic
networks, ranging from a 22 x 22 mm coverslip-size device that fits neatly into a six-well plate
format, a 48 mm-long device that maintains cell culture area but maximizes surface area of the
interface region, and an early prototype of a multiplexed device in which the inlets branch on-chip
to feed two adjacent channel networks (B) Device fabrication flow chart. Fabrication entails
sealing microfluidic devices with membranes, attaching chambers to lids, and then bonding
devices to chambers……………………………………………………………………………………56
Figure 3-2. Microfluidic Devices to Co-Culture Hypoxic and Normoxic Cells. (A) Microfluidic
devices contain two channels for gas flow that mimic infarct and remote zones in post-infarct
myocardium. Gases diffuse across a 100 µm membrane to reach cells cultured on its surface. (B)
Channel networks are DLP 3D-printed and replica molded in PDMS. (C) Complete assembly
involves bonding to a PDMS membrane, chamber and lid. Chambers multiplex the device into
hypoxia, normoxia, and co-culture regions. (D) An optical oxygen sensor, PtOEPK (red),

viii
demonstrates adjacent regions of 0 and 10% oxygen along the device from n=3 independent
trials. Schematic in A is Created in Biorender.com………………………………………………….63
Figure 3-3. Figure 2. Co-Culture Between Hypoxic and Normoxic Human Cardiac Fibroblasts for
4h. Immunofluorescence of (A-B) HIF-1α (red), DAPI (blue), (C-D) Ki-67 (green), (E-F) α-SMA
(red), actin (green), and DAPI (blue). Error bars depict mean with standard deviation from n=3
independent trials. *p<0.05, according to ordinary one-way ANOVA with Tukey’s multiple
comparisons test. Brightness and contrast are adjusted on all images.
Scale bars = 100 µm…………………………………………………………………………………...64
Figure 3-4. Co-Culture Between Hypoxic and Normoxic Cardiac Fibroblasts for 24h.
Immunofluorescence of (A-B) HIF-1α (red), DAPI (blue), (C-D) Ki-67 (green), (E-F) α-SMA (red),
actin (green), and DAPI (blue). Error bars depict mean with standard deviation from n=3
independent trials. *p<0.05 **p<0.01 ****p<0.0001, according to ordinary one-way ANOVA with
Tukey’s multiple comparisons test or the Kruskall-Wallis test. Brightness and contrast are
adjusted on all images. Scale bars = 100 µm……………………………..…………………………65
Figure 3-5. Oxygen-Dependent Changes in Gene Expression in Human Cardiac Fibroblasts. (A)
Principal Component Analysis of RNA-sequencing data. (B) Venn diagram of differentially
expressed genes in all conditions versus normoxia. (C-D) Volcano plots and (E) pathway analysis
dot plots for shared pathways between hypoxia versus normoxia and hypoxic co-culture versus
normoxia. (F) Normalized expression of genes related glycolysis and wound healing. Error bars
reflect mean with standard deviation of three pooled samples across n=6 independent trials.
***p<0.0005 and ****p<0.0001, according to ordinary one-way ANOVA with Tukey’s multiple
comparison’s test…………………………………………...……………………..……………………67
Figure 3-6. Crosstalk-Dependent Changes in Gene Expression in Human Cardiac Fibroblasts.
Volcano plots and pathway analysis of differentially expressed genes in (A) normoxic co-culture
versus normoxia or (B) hypoxic co-culture versus hypoxia. (C) Normalized counts of genes
significant to co-culture. (D) Pro-inflammatory concentrations of IFN-γ and TNF-α in cell media
isolated from hypoxia, hypoxia-normoxia co-culture, or normoxia regions. Error bars depict mean
with standard deviation from three pooled samples across n=6 independent trials. *p<0.05
**p<0.01, according to ordinary one-way ANOVA with Tukey’s multiple comparisons test……70
Figure 3-7. Cytokine Levels in Media Isolated from Hypoxic, Normoxic, or Co-Culture Fibroblasts.
(A) Pro-inflammatory and (B) anti-inflammatory cytokine levels in media isolated from uniform
hypoxia, uniform normoxia, or hypoxia-normoxia co-culture regions of the device after 24h using
the MSD Human V-Plex Pro-Inflammatory 10-Plex Assay Kit. Statistical testing was performed
according to ordinary one-way ANOVA with Tukey’s multiple comparisons test or the KruskalWallis test……………………………………………………………………………………………..….71
Figure 4-1. Microfluidic Devices Adapted to Co-Culture Hypoxic Cardiac Fibroblasts with
Normoxic Cardiac Myocytes. (A) Schematic of microfluidic devices, which contain two channels
for hypoxic and normoxic gas flow. Gases diffuse across the cell culture membrane to reach cells
cultured on the surface. Devices are bonded to a cell culture chamber which multiplexes the
device into a hypoxia, normoxia, and co-culture region. (B) Flow chart depicting use of a CoSeeding Chamber to enable seeding of two distinct cell types on microfluidic devices……..….82
Figure 4-2. Co-Culture of Normoxic hiPSC-derived Cardiac Myocytes and Hypoxic Cardiac
Fibroblasts for 24h. (A) Brightfield of cardiac fibroblasts (CFs) and myocytes (CMs) 12h after cell
seeding. Microfluidic channels can be visualized beneath the cell culture membrane. (B)

ix
Fluorescence microscopy of DAPI (blue), vimentin (green), and α-actinin (red) in hypoxic CFs and
normoxic CMs in control or co-culture regions for 24h. Brightness and contrast are adjusted on
all images. Scale bars = 500 µm. Quantification of nuclei (C) size and (D) number using the
ImageJ analyze particles function. Error bars depict mean with standard deviation from n=3
independent trials. (E) Total RNA concentration from samples used for RNA-sequencing. Error
bars depict mean with standard deviation from two pooled replicates across n=4 independent
trials. * p < 0.05, *** p < 0.0005, and **** p < 0.0001 according to ordinary one-way ANOVA with
Tukey’s multiple comparisons test……………………………………………………………………84
Figure 4-3. The Global Transcriptome of Cardiac Fibroblasts (CFs) and Cardiac Myocytes (CMs)
in Hypoxia, Normoxia, or Co-Culture for 24h. (A) Principal component analysis and (B)
hierarchical clustering on RNA-sequencing data demonstrates primary clustering of samples
based on oxygen-level and cell type, regardless of co-culture. Each sample reflects a pooling of
two samples, for a total of four independent trials…………………………………….…………….85
Figure 4-4. Crosstalk-Dependent Changes in Gene Expression in Hypoxic Human Cardiac
Fibroblasts and Normoxic hiPSC-derived Cardiac Myocytes Co-Cultured for 24h. (A) Volcano
plots and (B) pathway analysis of differentially expressed genes between the central co-culture
and control regions. (C) Dot plots for co-culture specific genes and their expression in hypoxic
cardiac fibroblasts (CFs) and normoxic cardiac myocytes (CMs) in co-culture versus control
regions. Error bars reflect mean with standard deviation. Each sample is a pooling of two samples,
for a total of four independent trials. * p < 0.05, ** p < 0.005, *** p < 0.0005, and **** p < 0.0001
according to ordinary one-way ANOVA with Tukey’s multiple comparisons test or the KruskalWallis test………………………………………………………………………………………………..88
Figure 4-5. Cytokine Secretion in Hypoxic Human Cardiac Fibroblasts and Normoxic hiPSCderived Cardiac Myocytes Co-Cultured for 24h. Mean concentrations of an array of ten (A) proand (B) anti-inflammatory cytokines in cell media. Error bars reflect mean with standard deviation
from two pooled samples across four independent trials. * p < 0.05, ** p < 0.005, *** p < 0.0005,
and **** p < 0.0001 according to ordinary one-way ANOVA with Tukey’s multiple comparisons
test or the Kruskal-Wallis test………………………………………………………………………….90

x
Abstract
A myocardial infarction, or heart attack, disrupts the flow of vital, oxygenated blood to
downstream tissue, leading to cardiac myocyte necrosis within hours. Because the mammalian
heart has a limited capacity to regenerate, the subsequent wound healing process relies on the
formation of scar tissue. Cardiac fibroblasts play a critical role in this process by activating into a
myofibroblast phenotype, which deposits matrix that ultimately forms a scar. Although this is
necessary to maintain tissue integrity, excessive fibrosis often occurs and is a common feature of
heart failure. Whether crosstalk between hypoxic, injured tissue and neighboring, healthy
myocardium may regulate mechanisms of fibrosis has been relatively unexplored. This is due
largely to shortcomings of existing research tools, which offer limited spatiotemporal control over
relevant biophysical features of the cell microenvironment. Organs on Chips have recently
emerged to recapitulate structural and functional features of human tissue, such as oxygen level.
Our objective was to develop a Post-Infarct Myocardium on a Chip for the co-culture of hypoxic
and normoxic cardiac cells. We first characterized human cardiac fibroblast phenotypes in acute
hypoxia. We then engineered and implemented a Post-Infarct Myocardium on a Chip for oxygen
control to identify crosstalk between hypoxic and normoxic fibroblasts. Lastly, we adapted it for
the heterotypic co-culture of hypoxic fibroblasts and normoxic cardiac myocytes. We find that
oxygen-dependent crosstalk exacerbates pro-inflammatory signaling. This work contributes to
Organ on Chip models of post-infarct myocardium and has applications in disease modeling and
drug screening.

1
Chapter 1 Introduction: Engineering the Cellular Microenvironment of Post-Infarct
Myocardium on a Chip
This Chapter is an adaptation of: Khalil Natalie N., and McCain Megan L. Engineering the Cellular
Microenvironment of Post-infarct Myocardium on a Chip. Frontiers in Cardiovascular Medicine.
2021.
Myocardial infarctions are one of the most common forms of cardiac injury and death
worldwide. Infarctions cause immediate necrosis in a localized region of the myocardium, which
is followed by a repair process with inflammatory, proliferative, and maturation phases. This repair
process culminates in the formation of scar tissue, which blocks synchronous contraction of the
heart and often leads to heart failure in the months or years after the initial injury. In each
reparative phase, the infarct microenvironment is characterized by distinct biochemical, physical,
and mechanical features, such as inflammatory cytokine production, localized hypoxia, and tissue
stiffening, which likely each contribute to physiological and pathological tissue remodeling by
mechanisms that are incompletely understood. Traditionally, simplified two-dimensional cell
culture systems or animal models have been implemented to elucidate basic pathophysiological
mechanisms or predict drug responses in cardiac cells exposed to select features of the infarct
microenvironment. However, these conventional approaches offer limited spatiotemporal control
over relevant features. To address these gaps, Organ on a Chip models of post-infarct
myocardium have recently emerged as new paradigms for dissecting the highly complex,
heterogeneous, and dynamic post-infarct microenvironment. In this Chapter, we first extensively
describe the cell and tissue-level remodeling that occurs following infarction to provide a basis for
microenvironmental features to model on a Chip. We then describe recent Organ on a Chip
models of post-infarct myocardium, including their limitations and future opportunities in disease
modeling and drug screening.

2
1.1. The Clinical Perspective
Cardiovascular disease is the leading cause of death worldwide, responsible for over 17.9
million deaths annually1
. A common cause of cardiovascular disease is coronary artery disease,
which refers to plaque buildup in the coronary arteries. Complete occlusion of the coronary
arteries can lead to a myocardial infarction, or heart attack2
. Myocardial infarctions are among the
top five most expensive conditions treated by US hospitals, and affect over 800,000 Americans
annually3, 4
.
1.1.1. Clinical Diagnoses
The evaluation of a myocardial infarction includes the assessment of presenting
symptoms, electrocardiographs, and cardiac troponin serum levels. Patients with symptoms of
myocardial infarction, for example chest pain caused by damage to the cardiac muscle,
immediately undergo electrocardiography to assess electrical activity of the heart.
Electrocardiographs with ST segment elevation are indicative of a transmural infarction and may
result in diagnosis for ST-segment elevation myocardial infarction (STEMI). Patients without STsegment elevation may still have elevated cardiac troponin levels in serum indicating myocardial
necrosis and therefore be admitted for non ST-segment elevation myocardial infarction
(NSTEMI)5
.
Cardiac troponins, specifically troponin T and troponin I, are sarcomeric proteins specific
to cardiac myocytes that regulate contraction and are released into circulation within 8 hours after
myocardial injury. Other biomarkers suggesting myocardial necrosis include creatine kinase and
myoglobin, although these are not specific to the myocardium. Lastly, there are emerging
diagnostic markers associated with inflammation and plaque destabilization6
.

3
1.1.2. Current Treatments
The rapid identification of myocardial infarction in patients is essential for timely
reperfusion of the occluded coronary artery, for example through percutaneous coronary
intervention, to minimize the initial insult and resulting infarct size7
. Percutaneous coronary
intervention refers to several nonsurgical techniques to restore flow to the interrupted artery. This
includes angioplasty, which refers to balloon inflation via catheterization, and may be performed
alone or with stent insertion to ensure the vessel stays open after the balloon is deflated and
withdrawn; atherectomy, in which plaque is scraped away using a blade connected to a catheter;
or thrombectomy, in which plaque is captured and removed8
. Immediate pharmacological
treatments may also be administered to break up or prevent the formation of blood clots, and
these include thrombolytics, anticoagulants, or antiplatelet drugs, such as aspirin, heparin9
, or
P2Y12 inhibitors10
.
The first one to three months following revascularization procedures is referred to as the
subacute period after myocardial infarction10. Several therapies have proven benefit in reducing
adverse cardiovascular events, such as stroke, recurrent infarctions, and death, in the subacute
period10. These include treatments to improve cardiac function, dilate blood vessels, reduce the
amount of work done by the heart, alleviate pain, and prevent arrythmias9
.
RAAS Inhibitors
The renin-angiotensin-aldosterone system is a paracrine/endocrine system that regulates
blood pressure and volume through vasoconstriction, water retention, hypertrophy, and fibrosis,
among other cardiovascular processes11. Angiotensin II, the main effector molecule of this
system, increases blood pressure through vasoconstriction, which requires increased work by the
heart to pump blood throughout the body. Therefore, inhibiting angiotensin signaling via
angiotensin-converting enzyme (ACE) inhibitors or though angiotensin receptor blockers (ARBs)

4
is proven to reduce mortality in patients who have had a myocardial infarction10
. Angiotensin is
also known to promote fibrosis and increase collagen production in cardiac fibroblasts through
angiotensin receptor-mediated upregulation of transforming growth factor-β1 (TGF-β1).
Therefore, inhibiting angiotensin signaling may alleviate fibrosis12
. Furthermore, aldosterone
blockers are diuretics that inhibit aldosterone-stimulated water reabsorption into blood vessels,
thereby reducing blood pressure and the amount of work done by the heart13. Aldosterone
antagonists initiated three days after myocardial infarction have been shown to reduce mortality
in patients with low left ventricular ejection fraction and signs of heart failure10
.
Beta Blockers
Beta adrenergic receptors are another therapeutic target activated by epinephrine or
norepinephrine to stimulate heart rate, strength of contraction, and overall cardiac output.
Inhibiting B1 adrenergic receptor-mediated signaling on cardiac myocytes through beta
blockers10, 12, such as bisoprolol, carvedilol, and metoprolol, can reduce heart rate in hypertensive
patients following myocardial infarction10. The American Heart Association recommends beta
blocker therapy within 24 hours of initial myocardial infarction, and beta blockers have been
shown to reduce mortality rate in patients with reduced left ventricular ejection fraction10
.
Interestingly, while the most common beta blockers are B1 receptor-selective, cardiac fibroblasts
are known to express B2 adrenergic receptors which are linked to increased fibroblast
proliferation and IL-6 production; inhibition of B2 receptors may also reduce fibrosis following an
infarction12
.
Statin Therapy
Statins are inhibitors of HMG-CoA reductase, an intermediate in the metabolic pathway
that produces cholesterol14. Statin therapy therefore plays a role in lowering cholesterol and can
be used to prevent primary or secondary myocardial infarction12. Statins have been shown to

5
reduce mortality and acute coronary syndrome in patients with cardiovascular disease10. Finally,
lifestyle changes following myocardial infarction are also associated with reduced mortality, and
include improved diet and exercise10
.
While existing pharmacological interventions can improve the risk of heart failure and
mortality, and prevent secondary infarction, they have been relatively ineffective in reducing the
initial infarct size and alleviating fibrosis. Furthermore, reperfusion through percutaneous
coronary intervention may even lead to reperfusion injury, which further contributes to cardiac
myocyte death and infarct size. As a result, there is a need for new therapies to recover cardiac
myocytes in the infarct zone and reduce scar size after infarction. Recent approaches under
clinical trials include ischemic post-conditioning to condition the heart with cycles of ischemia and
reperfusion, tissue revascularization to recover myocardial tissue, and cell-based therapies, such
as the transplantation of stem cells to the site of infarction in patients. The development of
successful clinical treatments for myocardial infarction may require new model systems that can
increase our understanding of the complex pathophysiological mechanisms involved in the
cellular and molecular response to infarction15
.
1.2. Healing After Myocardial Infarction
1.2.1. Tissue-Level Response: Ventricular Remodeling
The myocardium undergoes many structural and functional changes in response to
myocardial infarction or other injury, a process termed cardiac remodeling16. In response to the
immediate loss of contractile tissue, the myocardium exhibits reduced left ventricular ejection
fraction. To maintain stroke volume, several compensatory mechanisms are employed, including
increasing muscle mass through concentric hypertrophy17 or increasing chamber volume through
ventricular enlargement18
.

6
Infarct Expansion
Within hours after myocardial infarction, before collagen deposition and scar formation
can increase the tensile strength of the myocardium19, the initial loss of tissue at the infarct zone
makes the region more vulnerable to deformation19, 20. This may result in infarct expansion, which
is defined as thinning of the infarct zone caused by slippage between necrotic myocyte bundles19
and a subsequent dilation of the left ventricle19-21. Infarct expansion can contribute to the likelihood
of myocardial rupture and the development of heart failure19, 20
.
Concentric Hypertrophy
As a result of infarct expansion, diastolic and systolic wall stresses increase16, 22. To
compensate, the myocardium increases its muscle mass and wall thickness through cardiac
myocyte hypertrophy17 in which cell volume can increase up to 78%19. Increased wall thickness
and cell volume normalize the increased load17. Myocyte hypertrophy and lengthening result in
increased circumference of the left ventricle and a shift towards a spherical geometry18. Together,
increased circumference and sphericity contribute to ventricular enlargement16
.
Ventricular Enlargement
Ventricular enlargement may be compensatory18 in order to maintain stroke volume after
initial reduction in left ventricular ejection fraction19, 21. However, dilation followed by insufficient
hypertrophy can lead to a vicious cycle of progressive ventricular enlargement and dysfunction19
.
In humans, end-systolic and end-diastolic chamber volumes can progressively increase up to 1
year after initial myocardial infarction23. Patients with ventricular enlargement have increased
likelihood of mortality19, 24, and ventricular volume and mass, along with long-term left ventricular
ejection fraction recovery, are commonly used to assess cardiac remodeling in clinical trials and
practice18, 25
.

7
Ventricular Dysfunction
Together, changes in the structure of the left ventricle can ultimately lead to ventricular
dysfunction. Ventricular dysfunction is clinically defined as an increase of 25% or greater in the
left ventricular end diastolic volume or a reduction below 45% in the left ventricular ejection
fraction26
. The 5-year survival rate for patients with ventricular dysfunction is 50%20
. Pathological
fibrosis during the wound healing process can also block synchronous contraction of cardiac
myocytes and introduce ventricular arrythmias20, resulting in further decline in ventricular function.
1.2.2. Cellular and Molecular Response: Phases of Wound Healing
At the cellular and molecular level, myocardial infarction is immediately followed by the
death of approximately one billion cardiac cells12. Localized hypoxia in the ischemic region results
primarily in cardiac myocyte necrosis27, but also apoptosis and autophagy28. Because the
mammalian heart has a limited capacity for regeneration, the subsequent repair process to
maintain tissue integrity is dependent on the formation of a scar. Repair of infarcted tissue occurs
through three overlapping phases: the inflammatory phase, proliferative phase, and maturation
phase.
Inflammatory Phase
The inflammatory phase occurs in the first 1-3 days following a myocardial infarction.
Necrotic cardiac myocytes release their intracellular contents, including mitochondrial DNA, heat
shock proteins, ATP, and other danger-associated molecular patterns (DAMPs) that warn the
body of injury and activate pro-inflammatory signaling pathways in innate immune cells29
.
Inflammatory chemokine and cytokine gradients comprised of tumor necrosis factor-α (TNF), IL1β, and IL-6 promote leukocyte migration into the infarct zone. Macrophages and neutrophils that
are recruited to the injured tissue will clear dead cellular debris and damaged extracellular

8
matrix29; initially, macrophages acquire an inflammatory phenotype, with augmented TNF and
proteinase expression to exacerbate the inflammatory response28. Matrix metalloproteinases
function to digest fibrillar collagen and assist phagocytosis of damaged ECM. By day 7 after
necrosis, tissue inhibitors of metalloproteinases (TIMPs) are upregulated that conclude this
degradative phase29. Ultimately, as neutrophils undergo apoptosis, macrophages are directed
towards a resolving phenotype and begin secreting anti-inflammatory signals, such as TGF-β and
IL-10, that repress the inflammatory response and drive cardiac fibroblast activation in the
subsequent proliferative phase12
.
Proliferative Phase
Over the next few weeks, the proliferative phase occurs that is characterized by cardiac
fibroblast proliferation, migration into the site of injury, and differentiation into an activated
myofibroblast phenotype. Cardiac fibroblasts are the most abundant non-myocyte cell population
in the mammalian myocardium29, and after expansion in the proliferative phase, in combination
with cardiac myocyte necrosis, they become the most abundant cell type in the infarcted region12
.
Myofibroblasts express α-smooth muscle actin and non-muscle myosin which provide them with
the ability to generate force to migrate and facilitate wound contracture12. Myofibroblasts
upregulate angiotensin II, its type 1 receptor, and TGF-β, which are involved in autocrine signaling
that results in myofibroblast collagen synthesis and fibrotic remodeling27. In the proliferative
phase, myofibroblasts deposit collagen, fibronectin, and other matrix proteins to maintain tissue
integrity and prevent myocardial rupture.
Maturation Phase
The final phase involves scar maturation, which occurs on a timescale of weeks to months
and is marked by collagen turnover, the deactivation of myofibroblasts, and an increase in the

9
tensile strength of the injured myocardium12, 30. The depletion of growth factors necessary for
myofibroblast survival ultimately leads to myofibroblast dispersion through apoptosis12, 27
.
Myofibroblasts that persist into the maturation phase begin secreting type I collagen in place of
type III collagen, which is referred to as collaged turnover. Fibroblasts secrete procollagen, a
collagen precursor, which must be cleaved extracellularly by collagen proteinases before selfassembly into fibrils that will then bundle to form mature collagen fibers. Further enzymatic
crosslinking of collagen occurs through lysyl oxidases (LOX) and progressively increases the
tensile strength of the myocardium for months after myocardial infarction12
.
1.2.3. The Pathological Fibrotic Response
In both health and fibrotic disease, the extracellular matrix in the myocardium consists
primarily of type I fibrillar collagen, which has the tensile strength of steel. The matrix consists of
an endomysium, which encapsulates individual myocytes into muscle bundles, a perimysium,
which segregates muscle bundles into myofibers, and an epimysium, which clusters myofibers to
preserve their orientation. This enables myocyte alignment for optimal force generation and
provides the myocardium with strength to resist deformation, myocyte slippage, and myocardial
rupture27
.
While myofibroblasts play a vital role in reparative fibrosis to maintain structural integrity after
the initial loss of tissue, they may also be driven by many cellular and molecular events towards
a pathological remodeling process. For example, fibrogenic growth factors, myofibroblast
persistence in the myocardium, increased mechanical stress, and excessive synthesis or
insufficient degradation of matrix proteins may all lead to the continued activation of fibroblasts
and subsequent scar expansion into non-infarcted regions12
.
Myofibroblast dispersion through apoptosis normally occurs during infarct healing; however,
myofibroblast persistence in the myocardium can occur months, or even years, after injury and is

10
a common feature of heart failure27. Fibrotic tissue has been previously described as a “living”
rather than inert tissue, whose secretome – including TGF-β, angiotensin II and its receptor, and
matrix proteins – can traverse the interstitial space and promote fibrosis in non-infarcted regions27,
31
.
Pathological fibrosis has many functional consequences. First, cardiac myocytes become
stress shielded by encapsulating fibrillar collagen, which reduces their workload and may result
in disuse atrophy. Second, fibrosis can physically block synchronous conduction of cardiac
myocytes, and similarly, direct coupling between myocytes and myofibroblasts can result in
abnormal impulse conduction. Excessive collagen deposition, especially if it occurs in areas
remote to the infarction, is ultimately responsible for the development of lifelong arrythmias and
the transition to heart failure12, 27. The factors that drive pathological fibrosis are incompletely
understood, in part due to a lack of experimental models that capture early environmental
changes after myocardial infarction.
Post-infarct myocardium is therefore characterized by a variety of biochemical and
biomechanical properties changing in both space and time, which are correlated to complex
remodeling of several cell types and the extracellular matrix in parallel (Figure 1-1). Remodeling
ultimately impacts cardiac function and patient outcomes. However, the relationships between
biochemical, biomechanical, cellular, and molecular factors are incompletely understood,
hindering the discovery of new therapies to mitigate the effects of the initial injury. Thus, there is
a great need for controlled experimental models of post-infarct myocardium that account for
remodeling of the cellular microenvironment to uncover mechanisms of pathophysiology.

11
Figure 1-1. The post-infarct microenvironment is characterized by dynamic physical, mechanical,
and biochemical cues that orchestrate the three phases of healing. Regional and temporal
changes of oxygen concentration, stiffness, inflammatory cytokines, cell and matrix composition
characterize the myocardium as injured tissue is replaced with a scar.
Created with BioRender.com.
1.3. In Vitro Hypoxia Models
Due to the high metabolic activity of cardiac myocytes, the myocardium is a highly
vascularized tissue, with capillaries separated by approximately 20 microns32. This translates to
about one blood vessel between every two cardiac myocytes33. As a result, hypoxia is one of the
most injurious effects of a myocardial infarction. Conventionally, hypoxia has been induced by
culturing cells in environments with uniformly low oxygen. However, phosphorescent oxygen
probes have demonstrated that a spatial gradient ranging from 0% to 10% oxygen bridges injured
tissue with neighboring viable tissue in post-infarct myocardium34, 35. In addition, oxygen
concentrations change over time as the infarct zone is re-oxygenated during reperfusion. Thus,
in vitro systems that can modulate oxygen concentrations in space or time have more recently

12
been developed to mimic the hypoxic landscape of post-infarct myocardium. In this section, we
will describe conventional hypoxia models that replicate uniform hypoxia as well as engineered
systems that offer spatial or temporal control over oxygen tension.
1.3.1. Conventional Hypoxia Models
To recapitulate myocardial hypoxia in vitro36, one of the most common approaches is to place
cardiac cells in incubators or hypoxia chambers and replace oxygen with nitrogen. This enables
the stable, long-term induction of hypoxia, with tunable control over global oxygen levels by
selecting a gas composition of choice. To enable cell handling, hypoxia workstations have also
been developed that contain gloveboxes to allow for the manipulation of cells in a hypoxic
enclosure. However, these approaches for physical induction of hypoxia require access to
specialized equipment, such as incubators with oxygen regulation, and are limited to uniform gas
concentrations.
Hypoxia can also be simulated in cells cultured in ambient oxygen by adding hypoxia mimetic
agents, such as cobalt chloride, to cell media. Hypoxia mimetic agents often work by stabilizing
hypoxia inducible factors (HIF), a family of transcription factors that facilitates the cellular
response to hypoxia by upregulating genes associated with survival in low oxygen. In normoxia,
HIF is constantly degraded through hydroxylation of the HIF-α subunit by the enzyme prolyl
hydroxylase, which marks it for ubiquitination by Von Hippel Lindau protein and subsequent
degradation. Cobalt chloride chemically stabilizes HIF in normoxia by replacing Fe2+ with Co2+ in
the prolyl hydroxylase active site, thereby inhibiting its hydroxylation of HIF-α and the ensuing
degradation pathway. Other successful hypoxia mimetics include dimethyloxalylglycine and
deferoxamine, which similarly work by inhibiting prolyl hydroxylase activity37. Hypoxia mimetics
enable easy access to cells during cell culture, are inexpensive, and can quickly simulate hypoxic
conditions in vitro. However, they can have adverse effects on other signaling pathways that are

13
not affected by low oxygen37, 38, may be cytotoxic39, and likely do not capture all the effects of true
hypoxia.
Conventional in vitro models have shown that hypoxia results in structural and functional
changes that may contribute to the development of arrhythmias. Specifically, the gap junction
protein connexin 43 (Cx43) is known to be affected, which plays a critical role in conducting
electrical impulses in the myocardium. Hypoxia has been shown to drive electrical uncoupling in
vitro, with effects including decreased Cx43 signal at gap junctions40, 41, increased Cx43
internalization40 and dephosphorylation42, and decreased conduction velocity41. Hypoxia is also
associated with inactivation of sodium potassium pumps (Na,K ATPase), which regulate cardiac
action potentials43, 44. Hypoxic cardiac myocytes also upregulate fetal, T type calcium channels,
which are absent in healthy, adult myocardium, in a mechanism dependent on HIF-1𝛼
45. Finally,
hypoxia has also been shown to promote apoptosis in cardiac myocytes46-49
. These remodeling
processes are possibly related to the increased incidence of arrhythmias observed in post-infarct
myocardium.
Studies have also shown that cardiac fibroblasts and other non-myocyte cell populations are
generally more resistant to hypoxia46-48. In response to hypoxic stimuli in vitro, cardiac fibroblasts
undergo differentiation into the activated, myofibroblast phenotype, which is signified by increased
α-SMA expression50-52, collagen type I expression50-53, and migration capacity52. However, there
have been controversial reports of cardiac fibroblast proliferation in response to hypoxia50-52
,
which may result from differences in basal levels of fibroblast differentiation, or from differences
in cell source, oxygen concentration, and hypoxia duration. In response to prolonged hypoxia
exposure in vitro, cardiac fibroblasts ultimately undergo apoptosis54-56
.
Hypoxia also changes the secretion of paracrine factors that are involved in many processes
of infarct healing, such as angiogenesis, fibroblast differentiation, and remodeling of the
extracellular matrix. Hypoxic cardiac myocytes upregulate vascular endothelial growth factor
(VEGF)57, 58, insulin-like growth factor 258, inflammatory cytokines (TNF-α, IL-1β, IL-6), and TGF-

14
β
59. As TGF-β is known to promote fibroblast differentiation into myofibroblasts, medium
conditioned by hypoxic myocytes has been reported to drive cardiac fibroblast migration60 and
facilitate faster wound closure in cultured skin fibroblasts59. Similar to cardiac myocytes, hypoxic
cardiac fibroblasts upregulate TGF-β1 and also its receptor, TGFβ-R1, which can play a role in
autocrine signaling pathways to promote fibroblast differentiation52, 54. Hypoxic fibroblasts also
exhibit increased secretion of inflammatory cytokines (TNF-α
61-63 and IL-6
63), matrix
metalloproteinases (MMP-2 and MMP-9
62) and VEGF62, and conditioned medium from hypoxic
fibroblasts has been shown to reduce cardiac myocyte viability61, 64
. Thus, hypoxia likely alters
cellular cross-talk between distinct cardiac cell types in post-infarct myocardium.
1.3.2. Engineered Models: Spatially Controlled Oxygen
Although conventional hypoxia systems have revealed valuable insights into oxygendependent remodeling of cardiac cell types, they do not replicate the spatial or temporal changes
in oxygen that are characteristic of post-infarct myocardium. To mimic spatial oxygen gradients,
more complex in vitro systems have been engineered. In one example, a microfluidic device was
fabricated with a central channel designated for cell culture embedded between two lateral media
channels. By flowing media containing a chemical hypoxia mimetic through one channel and
standard cell medium through the other, a chemical hypoxia gradient was established in the
central cell-containing channel. Cardiac myoblasts near the hypoxic end of the gradient exhibited
altered morphology, including reduced cell area and actin disintegration, which was accompanied
by mitochondrial dysfunction and decreased cell viability65. In addition to chemical methods66-69
,
microfluidic devices have also been developed to generate physical oxygen gradients by culturing
cells on a gas-permeable membrane above microchannels for gas flow70-72. However, these
methods have not been extensively applied to cardiac cell types.

15
Ischemic gradients have also been developed by stacking thin layers of hydrogels that are
mechanically supported by a paper scaffold, a technique termed “cells-in-gels-in-paper.” To
control oxygen and nutrient diffusion into the stack, one end of the construct was placed in a base
that is impermeable to gases and liquids. Nutrients become depleted as they diffuse into the stack,
which creates an ischemic environment in the lower layers. Cardiac myocytes in the ischemic,
lower layers exhibited reduced viability and circular morphology when compared with upper
layers. Cell-tracking demonstrated that cardiac fibroblasts embedded in upper layers migrate
towards ischemic cardiac myocytes. Fibroblast migration increased when myocytes were
exposed to higher levels of ischemia, generated through taller stacks, and was reduced in the
absence of cardiac myocytes or with the pharmacological inhibition of TGF-β
32
. Other methods
have been developed that similarly modify hydrogels to generate oxygen gradients73 by using
oxygen-consuming enzymes during hydrogel cross-linking74, 75, embedding a hydrogel between
gas flow channels76, or linearly increasing cell density and thus oxygen consumption rates77, but
these methods have not yet been implemented to model post-infarct myocardium.
Lastly, cardiac spheroids have also been implemented to mimic hypoxia gradients.
Cardiac spheroids are 3-D aggregates of cardiac cells that recapitulate select aspects of native
tissue structure and function78-82. Because the diffusion limit of oxygen in tissues is around 100-
200 microns83, cardiac spheroids intrinsically generate oxygen gradients, for which oxygen
tension is highest at the surface and decreases towards the necrotic core. Though this is
conventionally thought of as a limitation of spheroids, recent work has harnessed this property to
develop ‘infarct spheroids’ that are exposed to ambient 10% oxygen to generate spatial hypoxia
gradients (0-10% oxygen) that mimic the infarct, border, and remote zones after infarction (Figure
1-2A). These infarct spheroids contained cardiac myocytes, endothelial cells, and stromal cells
and were treated with noradrenaline to mimic neurohormonal stimulation after infarction. Infarct
spheroids demonstrated similar global gene-expression profiles to human ischemic
cardiomyopathy and animal myocardial infarction samples. Furthermore, when compared with

16
control spheroids in ambient oxygen, infarct spheroids exhibited a metabolic shift towards
glycolysis, increased stiffness, increased expression of myofibroblast markers, decreased cardiac
myocyte contraction amplitude (Figure 1-2A), and asynchronization of contractions84
.
1.3.3. Engineered Models: Temporally Controlled Oxygen
In vitro models have also been engineered to replicate dynamic changes in oxygen
characteristic of ischemia-reperfusion. Microfluidic devices are particularly suitable for this
application because they contain chambers of small volumes that can be rapidly filled with hypoxic
gas or cell medium. For example, a microfluidic device with integrated bioelectronics was used to
measure intracellular action potential and extracellular beat rate and propagation velocity in
cardiac myocytes cultured in a microchannel. The microchannel was rapidly filled with hypoxic
cell medium followed by recovery medium to mimic ischemia-reperfusion. Hypoxic cardiac
myocytes demonstrated substantially reduced depolarization times and beat rates, as well as
irregular propagation patterns, which recovered within 30 minutes after reintroducing normoxic
cell media85. A similar microfluidic device was fabricated to contain small chamber volumes that
can be quickly filled with hypoxic gas. Gas from the upper chamber diffused across a thin, gaspermeable membrane to reach cells cultured in lower microfluidic channels. This device
demonstrated that hypoxic conditions below 5% oxygen induce changes in cardiac myocyte
calcium transients, including a decrease in amplitude that could be mimicked using L-type calcium
channel antagonists. After a subsequent ten minutes of reperfusion with normoxic gas, cardiac
myocytes recovered with normal calcium transients86. Together, these studies suggest that
hypoxia induces reversible alterations in cardiac myocyte electrophysiology.
In another model of ischemia-reperfusion, cardiac myocytes were cultured as an aligned
tissue on silicone cantilever substrates with embedded strain sensors (Figure 1-2B). The cardiac
myocytes were aligned using microcontact printing, which involves the preparation of silicone

17
“stamps” that are used to transfer extracellular matrix proteins to a substrate in desired
geometries. With a spatial resolution of approximately 1 µm, microcontact printing can direct
tissue orientation87-89, single-cell shape90, and even subcellular structures91
. When the aligned
cardiac tissues contracted on the cantilever substrates, the cantilevers deflected, leading to a
resistance change in the embedded strain sensors proportional to the contractile stress92, 93. This
system was used to provide real-time measurements of contractile stress in a simulated ischemiareperfusion injury by switching cells from ischemic media at 1% oxygen to standard media at
ambient 21% oxygen. Integrated sensor readouts demonstrated that cardiac tissues stopped
contracting during ischemia and displayed a minor recovery of twitch stress during reperfusion.
In contrast, pre-treatment with endothelial cell-derived extracellular vesicles was cardioprotective
and enabled cardiac tissues to continue contracting during simulated ischemia and exhibit higher
recovery of twitch stress after reperfusion (Figure 1-2B)15
.
In summary, several technologies, including microfluidic devices, hydrogels, spheroids,
and strain sensors, have been implemented to mimic the spatial and temporal oxygen gradients
characteristic of post-infarct myocardium and subsequently quantify changes in cardiac cell
phenotypes. Unlike conventional systems, these approaches can be used to explore oxygendependent regional and temporal changes in cellular phenotypes at high resolution and uncover
crosstalk between cells in distinct oxygen environments to identify new mechanisms of infarct
remodeling.

18
Figure 1-2. Engineered in vitro hypoxia models that modulate oxygen in (A) space and (B) time.
(A) Infarct spheroids generate intrinsic spatial oxygen gradients to model the infarct, border, and
remote zones after myocardial infarction. Infarct spheroids exhibit higher levels of myofibroblast
marker expression, exhibit asynchronous contraction of myocytes, and have lower contraction
amplitudes than control spheroids. Reprinted by permission from Springer Nature: Nat. Biomed.
Eng., (Richards et al., 2020). (B) Muscular thin films with integrated strain gauges can measure
changes in contractility in cardiac myocytes pre-treated with endothelial extracellular vesicles in
a simulated ischemia-reperfusion injury. Endothelial extracellular vesicles are cardioprotective
and provide cardiac tissues with enhanced contractility during both ischemia and reperfusion.
From (Yadid et al., 2020). Reprinted with permission from AAAS.
1.4. In Vitro Stiffness Models
Healthy myocardium is a moderately stiff tissue, with an elastic modulus of around 10 kPa.
After myocardial infarction, local elastic modulus in the infarcted region increases to 20-100 kPa
due to scar formation and fibrosis94-97. Rat models with coronary artery ligation demonstrate
myocardial stiffening over time, with elastic modulus increasing from 18 kPa to 55 kPa96, and
increased stiffness has been observed in the infarct zone as early as one day post-infarction98. In
addition to temporal changes, regional stiffness varies between the infarct zone, border zone, and
remote zone by three days post-infarction99. Elastic modulus progressively decreases in the
border zone at a rate of 8.5 kPa/mm towards remote tissue96
. Because investigating the effects
of tissue stiffness with in vivo models is confounded by many other concurrent changes, including

19
matrix composition98
, in vitro models have been implemented to elucidate the effects of stiffness
and other aspects of fibrosis on cardiac cell phenotypes. In this section, we will describe 2-D and
3-D in vitro models of cardiac fibrosis that focus primarily on recapitulating uniform or
spatiotemporal changes in stiffness.
1.4.1. Two-Dimensional Models with Uniform Stiffness
Because standard polystyrene dishes used for cell culture are nearly five orders of
magnitude more stiff than the native myocardium100, mechanically tunable biomaterials have been
developed to mimic the rigidity of healthy or fibrotic myocardium101
. For example, hydrogels are
cross-linked, hydrophilic polymers with high water content that are commonly used as cell culture
substrates because they can be tuned to resemble the elasticity of soft tissue and allow for
efficient mass transfer. Hydrogels can incorporate natural polymer chains, such as mammalian
matrix proteins, or synthetic polymer chains, such as polyacrylamide or polyethylene glycol101-103
.
Other biomaterials commonly used to recapitulate physiological or pathological stiffness in vitro
include elastomers like polydimethylsiloxane (PDMS), which is biocompatible, mechanically
tunable, and transparent104-107
.
When cultured on rigid hydrogel or elastomer substrates, cardiac myocytes exhibit
disorganized sarcomeres, reduced sarcoplasmic calcium stores108, lower amplitude calcium
currents108, 109, decreased cell shortening during contraction108, and a progressive decrease in
beat frequency over time110 when compared to substrates that mimic the elasticity of healthy
myocardium. Using traction force microscopy, in which fluorescent beads embedded in substrates
are displaced during cell contraction, several studies have established non-monotonic
relationships between force generation and substrate rigidity. In these studies, cardiac myocytes
generally generate maximum forces on physiological stiffness, which decrease on substrates that
are either more soft or stiff in both isotropic108, 111 and aligned microtissues112 . However, some

20
studies have reported linearly increasing force generation with increased substrate stiffness107,
113, 114, which may be attributed to differences in experimental methods, such as cell source,
biomaterial substrate, or analysis techniques. Similar results are observed in cocultures of cardiac
myocytes and fibroblasts on polyacrylamide substrates, in which increased stiffness results in
reduced troponin I staining, increased fibroblast density, and poor electrical excitability111
.
Micropatterning has also been used to modulate both cellular architecture and substrate
stiffness because both of these features remodel concurrently in post-infarct myocardium.
Substrate stiffness and cellular architecture has been shown to modulate metabolic activity104, 106,
112 and mitochondrial structure in cardiac myocytes105
. Microcontact printed hydrogels have also
been used to characterize the contractility of single90 or coupled115 cardiac myocytes as a function
of both cellular architecture and substrate stiffness. At the single cell level, cardiac myocytes with
low cell aspect ratios that mimic concentric hypertrophy do more work on stiff substrates that
resemble fibrotic myocardium, demonstrating a functional advantage of cell shape remodeling in
response to mechanical overload90. In coupled myocytes, stiff substrates caused increased focal
adhesion formation at the cell-cell interface, possibly contributing to cellular uncoupling in postinfarct myocardium115
.
To model both the cellular and biomechanical aspects of fibrosis, tissues have also been
engineered with cardiac myocytes and fibroblasts on substrates with tunable stiffness. For
example, microcontact printing has been implemented to engineer aligned microtissues on
polyacrylamide hydrogels with both cardiac myocytes and fibroblasts. Microtissues generated
less work on rigid substrates, irrespective of cell adhesion ligand or presence of fibroblasts,
revealing the dominant role of substrate elasticity in regulating contractile output116. To engineer
an artificial infarct boundary, cardiac myocytes and fibroblasts have been cocultured on separate
halves of cell culture substrates with rigidities that range from healthy, 1-week post-infarct, and 2-
to 6-weeks post-infarct myocardium. The presence of cardiac fibroblasts in this coculture setting

21
attenuated mechanical signal propagation across the infarct boundary in a stiffness-dependent
manner117
.
In addition to affecting cardiac myocytes, rigid substrates that mimic fibrotic myocardium
also promote fibroblast activation to myofibroblasts. On stiff substrates, cardiac fibroblasts notably
activate into a myofibroblast phenotype, exhibiting increased α-SMA coverage118-121, increased
contractile force generation measured through traction force microscopy122, and increased
nuclear localization of the mechanosensitive transcription factors yes-associated protein (YAP)
and transcriptional co-activator with PDZ-binding motif (TAZ)120. Knockdown of YAP and TAZ
reversed or attenuated stiffness-dependent changes in cell morphology and function, suggesting
YAP and TAZ coordinate fibroblast mechanoactivation120. In other work, limiting focal adhesion
size through microcontact printing was also sufficient to interrupt the recruitment of α-SMA to
stress fibers on stiff substrates, indicating that focal adhesion size may control α-SMA
localization119. Studies that establish mechanisms behind fibroblast mechanoactivation may
reveal new targets for anti-fibrotic strategies to mitigate adverse remodeling following myocardial
infarction.
In vitro models have also identified stiffness-dependent secretion of paracrine factors,
which may regulate several processes involved in infarct remodeling. For example, cardiac
myocytes on stiff substrates secrete more VEGF. Consistent with this finding, media conditioned
by myocytes on stiff substrates promotes angiogenesis, including increased migration capacity
and tube length of microvascular endothelial cells123. Fibroblasts cultured on stiff hydrogels and
treated with TGF-β have also been shown to upregulate several cytokines, including osteopontin,
a known regulator of collagen cross-linking via lysyl oxidase, and insulin-like growth factor 1,
which regulates cardiac myocyte hypertrophy. As a result, conditioned media from TGF-β-treated
fibroblasts cultured on stiff hydrogels has been shown to induce cardiac myocyte hypertrophy,
indicated by increased cell volume, when compared with medium from cardiac fibroblasts without
TGF-β, regardless of substrate stiffness124. This echoes previous work that demonstrates TGF-β

22
may be a more potent regulator of the myofibroblast phenotype than substrate rigidity118
.
Together, 2-D models that resemble the elasticity of fibrotic myocardium recapitulate many
cellular and molecular events following myocardial infarction.
1.4.2. Two-Dimensional Models with Spatiotemporal Control over Stiffness
2-D models with uniform stiffness do not encompass regional changes in stiffness
between the infarct, border, or remote zones following myocardial infarction. Spatial stiffness
gradients have been fabricated through graded material cross-linking125, including gradientpatterned126-129 or sliding130, 131 photomasks, layering of hydrogels of different elasticities132
,
applying a temperature gradient to PDMS during curing133, or microfluidic-mixing of prepolymer
solutions with different cross-linker concentrations134. However, most of these have not been
implemented with cardiac cell types to model post-infarct myocardium. In one example, a
polyethylene glycol hydrogel was patterned with soft and stiff concentric circles using a
photomask. Cardiac fibroblasts cultured on stiff regions of the substrate expressed increased αSMA and collagen when compared with soft regions. Live imaging demonstrated a directional
cellular migration towards the inner stiff region. Lastly, treatment with a ROCK inhibitor reduced
the population of myofibroblasts, demonstrating that the model can be used as an antifibrotic drug
screening platform135. To investigate the effects of pathological matrix stiffening in lung fibroblasts,
a stiffness gradient was made from polyacrylamide gels polymerized through gradient
photomasks. Human lung fibroblasts cultured on the stiffness gradient show a progressive
increase in fibroblast activation, indicated by increased proliferation and matrix synthesis, towards
the stiff end of the gradient. Addition of prostaglandin E2, an inhibitor of fibrogenesis, inhibited
fibroblast activation136. Similar phenotypes may also be observed in cardiac fibroblasts over
stiffness gradients, though this has not yet been tested.

23
Mechanical properties can also be controlled in situ to model changes in stiffness over
time, which is characteristic of infarct scar maturation. This can be achieved with materials that
polymerize in response to light exposure137, 138 or by varying the molecular weight of the crosslinking agent in real-time139. Engineered models to capture dynamic stiffening have been used to
model development, wound healing, and disease140, but few have been implemented in the
context of myocardial infarction. In one study, hyaluronic acid hydrogels seeded with cardiac
fibroblasts were modified to dynamically increase in stiffness in response to UV exposure (Figure
1-3A). Dynamic stiffening to model scar maturation resulted in increased cell spreading, α-SMA
formation, and collagen I expression (Figure 1-3A)121
. Although fibroblast activation correlates
with increased stiffness in 2-D spatiotemporal models, cardiac myocyte phenotype and function
has been relatively unexplored in these settings.
1.4.3. Three-Dimensional Models with Uniform Stiffness
Stiffness in 2-D only exposes one side of cells to the fibrotic microenvironments
experienced in vivo. To more closely mimic cell-cell and cell-matrix interactions that occur in native
myocardium, cardiac cells have been mixed into hydrogels to form 3-D tissues. Similar to findings
in 2-D, cardiac myocytes encapsulated in rigid polyethylene glycol hydrogels demonstrate
reduced cell shortening and increased relaxation time when compared with soft hydrogels, which
was also accompanied by increased intracellular localization of the mechanosensitive
transcription factor YAP138. In 3-D, matrix stiffness promotes fibroblast differentiation into
myofibroblasts, demonstrated by increased stellate morphology, α-SMA and collagen type III
levels, and gel compaction141, consistent with findings in 2-D. The simple aggregation of cardiac
fibroblasts in 3-D using low-attachment plates has also been shown to induce gene expression
changes associated with adverse cardiac remodeling and the extracellular matrix. Conditioned
media from 3-D fibroblast aggregates also causes cardiac myocyte hypertrophy relative to media

24
from fibroblasts cultured in 2-D142, indicating that phenotypes in 2-D do not always translate to 3-
D.
Using microfabricated templates, 3-D cardiac tissues have also been engineered with
control over cell composition, matrix stiffness, and tissue architecture. In one model, cardiac
myocytes and fibroblasts were embedded in collagen hydrogels of varying fibroblast cell densities
or collagen concentrations and suspended between uniaxial PDMS microposts. Microposts serve
as tissue constraints that promote alignment. Increasing fibroblast density decreased tissue
contraction force and hampered beating frequency, as measured by displacement of the
microposts143. In a similar paper, the system was modified to contain biaxial PDMS microposts to
generate isotropic cardiac matrices, designed to mimic ‘diseased’ architecture. 3-D microtissues
of cardiac myocytes and fibroblasts in isotropic matrices display more stellate morphology,
characteristic of myofibroblasts, and more heterogeneous force distribution when compared with
‘healthy’ aligned matrices. Furthermore, increasing the proportion of fibroblasts in the tissues
reduces the overall tissue beating frequency, suggesting that both matrix organization and cellular
composition regulate cardiac function144
.
Although hydrogels are mechanically tunable, they fail to recapitulate the fibrous
architecture of native cardiac extracellular matrix. A 3-D fibrous network functionalized with
fibronectin, which anchors cardiac cells in vivo, was fabricated through electrospinning. Spin
speed was adjusted to tune fiber alignment while photo-initiated cross-linking was used to tune
fiber stiffness to mimic physiologic (9-14 kPa) or pathophysiologic (>20 kPa) tissues. Cardiac
myocytes in stiff, fibrous networks exhibit slower calcium flux, indicated by increased decay time
and increased peak-to-peak irregularity145. In another example, fibrous scaffolds with varying fiber
stiffness were fabricated through two-photon polymerization and seeded with cardiac myocytes
that lack expression of cardiac myosin binding protein C, which is thought to play a role in
sarcomere sliding during contraction. Mutations in this protein are also associated with
hypertrophic and dilated cardiomyopathy. While control cells were able to adapt to the increased

25
load with increasing contraction force, cells with the mutation displayed impaired contraction on
stiffer fibers. This work demonstrates the combined effects of mechanical stress and genetic
factors on contraction deficits146
.
Interestingly, fibroblasts in 3-D fibrous matrices depart from the conventional relationships
established between stiffness and fibroblast activation in 2-D cell culture or 3-D hydrogels147. In
human lung fibroblasts seeded in the same fibrous matrices, increasing fiber stiffness actually
reduced proliferation and myofibroblast activation (α-SMA) when compared with cells on soft and
deformable fibrous matrices. This is correlated with reduced cell spreading and focal adhesion
formation that was also observed with increasing stiffness148. Fiber density, on the other hand,
has been shown to promote differentiation in lung fibroblasts, signified by increased fibronectin
synthesis, nuclear localization of YAP, proliferation, and cytokine secretion149. Similar
relationships may also exist for cardiac fibroblasts but have yet to be investigated.
1.4.4. Three-Dimensional Models with Spatiotemporal Control over Stiffness
Engineered 3-D cardiac tissues have also been fabricated with increasing spatiotemporal
control over stiffness. A 3-D fibrosis model was developed using the biowire platform, in which
cardiac cells are encapsulated in a fibrin-based hydrogel and suspended between a pair of
polymer wires that function to promote microtissue alignment. Tissue contractile stress is
measured based on the deflection of the intrinsically fluorescent polymer wires. To model healthy
(7.6 kPa) or fibrotic (61.1 kPa) myocardium, cardiac myocytes were cocultured with 25 or 75%
cardiac fibroblasts, respectively. Fibrotic tissues underwent more rapid compaction and had
higher collagen content, disrupted myofibril structures, altered Cx43 distribution, prolonged time
to peak, and lower contractile force generation when compared with healthy tissues. To next
create a spatially heterogeneous stiffness model, which can mimic the interface between the
infarct zone and viable tissue, both fibrotic and healthy tissue were integrated at opposing sides

26
of a single biowire platform (Figure 1-3B). The fibrotic side of the microtissue underwent more
rapid compaction, contained increased collagen content and myofibroblast activation (Figure 1-
3B), and had slower calcium transients with a lower amplitude compared to the healthy side. In
addition, propagation velocity at the healthy side was diminished when compared with uniform
healthy biowire tissues. Arrhythmic waves were also observed, especially in the interface region.
This platform was also used to screen antifibrotic drugs150, demonstrating the potential impact of
these approaches in drug development.
To alter substrate rigidity over time, one approach is to encapsulate cells into hydrogels
that either degrade or cross-link in response to specific wavelengths of light. In one example, a
3-D photodegradable hydrogel was used to demonstrate that valvular myofibroblasts can be
redirected into a quiescent phenotype by decreasing stiffness. This work demonstrates fibroblast
phenotypic plasticity and the potential role of the mechanical environment in de-differentiating
fibroblasts, which has therapeutic applications in resolving fibrotic disease151. Lastly, one study
demonstrated that cardiac myocytes encapsulated in photopolymerizable polyethylene glycol
hydrogels do not exhibit differences in cell viability after UV exposure138, though the impact of
progressive stiffening in 3-D on cardiac cell phenotypes has not been further established.

27
Figure 1-3. Engineered in vitro fibrosis models that modulate stiffness in (A) time and (B) space.
(A) A hyaluronic acid hydrogel is polymerized in situ upon exposure to UV light to mimic scar
maturation after myocardial infarction. Stiffening from 8 to 30 kPa results in collagen turnover.
Republished with permission of the American Society for Cell Biology, from (Herum et al., 2017a);
permission conveyed through Copyright Clearance Center, Inc. (B) A heteropolar biowire
integrates normal (7.6 kPa) and fibrotic (61.1 kPa) cardiac tissue. The white dashed line indicates
the interface between normal and fibrotic tissue. Normal regions exhibit increased α-actinin, a
component of sarcomeres in cardiac myocytes, while fibrotic regions have increased collagen
and α-smooth muscle actin, markers of myofibroblasts. Images reproduced with permission of the
American Chemical Society, from https://pubs.acs.org/doi/full/10.1021/acscentsci.9b00052
(Wang et al., 2019). Further permissions related to the material excerpted should be directed to
the American Chemical Society.
1.5. In Vitro Models of Pathological Strain
Cardiac cells are constantly under cyclic stretch in the healthy, beating heart. Myocardial
infarction results in an initial loss of contractility in the infarct zone followed by arrhythmogenesis,
which alters strain rates experienced by surviving cardiac cells, as quantified through
echocardiographic imaging152. To stretch myocytes and nonmyocytes in vitro, experimental
platforms include microchips with stretchable silicone membranes, custom-built bioreactors, or
commercially available cell straining units in which strain can be applied to cell culture plates with
integrated loading posts.
Chronic cyclic stretch over several days to mimic the diastolic and systolic movement of
cardiac muscle has been shown to promote the maturation of ‘engineered heart tissues’, which

28
are generally defined as cardiac myocytes embedded in hydrogels and cast around uniaxial tissue
constraints or circular molds. Stretched heart tissues exhibit increased cell alignment153-156
,
sarcomere organization156, 157, Cx43 expression155, 156, and contractile force generation153, 155, 158
.
In some studies, morphological changes were also observed that indicate cardiac myocyte
hypertrophy through increased cell size153, 157 and mitochondrial density153. In 2-D aligned cardiac
tissues fabricated through microcontact printing, chronic cyclic stretch has also been shown to
induce pathological changes in cell aspect ratio and sarcomere alignment, promote gene
expression profiles associated with pathological remodeling, and diminish calcium transients and
force generation (Figure 1-4A)159. Thus, chronic cyclic stretch can be beneficial or detrimental to
cardiac myocytes, depending on the specific parameters.
Cardiac fibroblast responses to strain have also been relatively inconsistent. In some
cases, stretching activates many hallmarks of cardiac fibrosis, including increased fibroblast
proliferation, hydrogel stiffening160, increased gel compaction and strength161, 162, extracellular
matrix deposition160-162, and enhanced secretion of TNF-α
163
. However, responses are dependent
on baseline levels of fibroblast activation, which is highly sensitive to culture conditions. Cardiac
fibroblasts that are cultured for one day on rigid substrates and have lower initial levels of α-SMA
respond to static tensile forces with increased α-SMA, while cells cultured for three days with
higher basal levels of α-SMA respond to the same force with decreased α-SMA production164
.
Consistent with this, fibroblasts grown on soft hydrogels with minimal basal α-SMA expression
and exposed to static stretch show elevated α-SMA mRNA levels and expression of various
extracellular matrix proteins, including collagen and fibronectin121. Fibroblasts also show
controversial proliferative behavior in response to mechanical strain, which may be dependent on
baseline α-SMA levels, strain rate165, ECM composition166, 167, serum concentration168, or
substrate stiffness121, which highlights a need for more standardized cell culture methods169
.
To assess the combined effects of strain and hypoxia, cardiac fibroblasts have been
cultured in a microfluidic device containing a stretchable, gas-permeable membrane situated

29
above a microchannel for gas flow and between lateral actuation channels (Figure 1-4B). Uniform
hypoxia (1% oxygen) or reduced contractility to mimic post-infarct myocardium (2% strain) are
alone sufficient to induce proliferation and collagen type 1 production, although the combined
effects of hypoxia and reduced strain are required to trigger fibroblast secretion of IL-1β or TGFβ (Figure 1-4B)170
.
Paracrine signals secreted by stretched cardiac cells may also regulate critical aspects of
infarct healing. Recent work has characterized the transcriptomic profile of stretched cardiac
myocytes, which show differentially expressed genes and regulatory networks that may lead to
hypertrophic growth of cardiac myocytes171. Consistent with this, stretched cardiac myocytes
upregulate miR208, a mediator of cardiac hypertrophy, hypertrophic proteins, such as β-myosin
heavy chain, and secretion of TGF-β
172
. Neonatal rat cardiac myocytes on stretched silicone
membranes have also been reported undergo apoptosis, accompanied by mitochondrial
dysfunction173
. One study explored paracrine signals secreted by stretched cardiac myocytes by
fabricating a coculture device that enables paracrine signaling between cardiac myocytes and
fibroblasts while exposing cardiac myocytes to strains that mimic the border zone after infarction
in vivo. In this device, coculture with stretched cardiac myocytes increases cardiac fibroblast
proliferation. A media screen indicated the presence of cytokines such as colony stimulating factor
1 and platelet derived growth factor B, which were sufficient to increase proliferation in fibroblast
monocultures121
.

30
Figure 1-4. Engineered in vitro models to mimic pathological strain post-infarction. (A) Muscular
thin films on a stretchable silicone membrane provide real-time measurements of contractile
stress during stretch. Images reproduced from (McCain et al., 2013). (B) A microfluidic device to
evaluate the combined effects of hypoxia (yellow channel) and strain (red channel) on cardiac
fibroblasts. Created with BioRender.com. This work demonstrates that the combined effects of
hypoxia (1% O2) and reduced contractility (2% strain) to mimic injured myocardium are required
for cardiac fibroblasts to upregulate TGF-β and IL-1β. Image reproduced from (Ugolini et al.,
2017). (https://creativecommons.org/licenses/by/4.0/).

31
1.6. Objective and Aims
As described above, innovative approaches to model the heterogeneous and dynamic
cellular microenvironments of post-infarct myocardium in vitro have identified new oxygen,
stiffness, and strain-dependent mechanisms that underlie pathological remodeling of many
cardiac cell types, which can manifest as hypertrophy, arrhythmias, and/or fibrosis. These
technological advancements will help to identify new therapeutic targets for myocardial infarction
and expedite the drug discovery pipeline. However, there are still many gaps in our understanding
of cardiac cell responses to the spatiotemporal complexities of post-infarct myocardium. As a
result, pathological remodeling after myocardial infarction continues to be a major clinical
problem.
Because the wound healing process involves a dynamic interplay between hypoxic,
injured tissue, and neighboring healthy myocardium, we hypothesized that crosstalk between
hypoxic and normoxic cardiac cells may regulate mechanisms of pathological fibrosis. To
investigate, our objective was to develop and implement an Organ on a Chip model of post-infarct
myocardium with spatiotemporal control over oxygen level. Because cardiac fibroblasts play a
critical role in driving pathological remodeling, in Chapter 2, we first aim to characterize human
adult cardiac fibroblast phenotypes in response to hypoxia on acute timescales. In Chapter 3, we
harness 3D-printing to rapidly prototype microfluidic devices for the real-time co-culture of hypoxic
and normoxic cardiac fibroblasts. Finally, in Chapter 4, we adapt this cell culture system to coculture hypoxic cardiac fibroblasts with normoxic human induced pluripotent stem cell (hiPSC)-
derived cardiac myocytes. Throughout these aims, we robustly characterize phenotype through
immunofluorescence, proteomic analysis, and transcriptomic analysis of cells, as well as
immunoassays of cell media. Together, this work advances existing Organ on a Chip models of
post-infarct myocardium and sheds light on the oxygen-dependent regulation of inflammation and
fibrosis following infarction.

32
Chapter 2 Human Cardiac Fibroblast Phenotypes Induced by Acute Hypoxia
This Chapter is an adaptation of: Khalil Natalie N., Rexius-Hall Megan L., Escopete Sean, Parker
Sarah J., McCain Megan L. Distinct phenotypes induced by acute hypoxia and TGF-β1 in human
adult cardiac fibroblasts. The Journal of Molecular and Cellular Cardiology Plus. In Review.
Myocardial infarction causes hypoxic injury to downstream myocardial tissue, which then
initiates a wound healing response that replaces injured myocardial tissue with a scar. Wound
healing is a complex process that consists of multiple phases, in which many different stimuli
changing in both space and time ultimately activate cardiac fibroblasts to differentiate into
myofibroblasts and deposit matrix. While this process is necessary to replace necrotic tissue,
excessive and unresolved fibrosis is common post-infarct and correlated with heart failure.
Therefore, defining the temporal dynamics of cardiac fibroblast phenotypes in hypoxia is essential
for understanding and ultimately mitigating pathological fibrosis. In this study, we treated primary
human adult cardiac fibroblasts with acute durations of hypoxia and then characterized their
phenotype through immunofluorescence, quantitative RT-PCR, and proteomic analysis. We
found that fibroblasts responded to low oxygen with increased localization of hypoxia inducible
factor 1 (HIF-1) to the nuclei after 4h, which was followed by increased gene and protein levels of
VEGFA and LOX, respectively. Hypoxia also inhibited proliferation after 24h, perhaps to conserve
energy. Consistent with this, we observed a global reduction in protein synthesis, including
collagen biosynthesis. In contrast, TGF-β1 treatment, a positive control for activation into the
myofibroblast phenotype, upregulated various fibrotic pathways. Collectively, these data suggest
that cardiac fibroblasts exhibit HIF nuclear localization, decreased proliferation, and a shut-down
in protein synthesis on acute timescales. Discerning the outcomes of temporal changes in hypoxia
treatment is important for elucidating the role of oxygen in fibrotic remodeling post-MI.

33
2.1. Introduction
Within one month of myocardial infarction, patients face a mortality rate of nearly 40%174
and another 13% develop heart failure175. Infarctions deprive myocardial tissue of oxygenated
blood, which initiates progressive necrosis of cardiac cells within 20 minutes176. This activates a
complex wound healing process that ultimately results in scar formation28. In this process, antiinflammatory mediators, such as TGF-β, drive activation of cardiac fibroblasts into a myofibroblast
phenotype28. Myofibroblasts are responsible for synthesizing and depositing matrix proteins that
form a scar12, 177, 178. They are also characterized by expression of alpha-smooth muscle actin (αSMA), an actin isoform that enables them to generate more force for wound contracture179, 180
.
Myofibroblasts should eventually disperse through apoptosis after wound healing, but they can
persist in the myocardium for months to years after injury27
. Excessive scarring by myofibroblasts
causes thickening and stiffening of the left ventricular wall and is a major cause of heart failure in
patients who have suffered myocardial infarction12
. A comprehensive understanding of
microenvironmental regulation of fibroblast phenotypes is critical to mitigating pathological fibrotic
remodeling post-infarction.
One of the key features of the post-infarct microenvironment is acute hypoxia. Studies
have shown that neonatal mouse and rat cardiac fibroblasts exposed to 1-2% O2 for 12-48 hours
increased expression of α-SMA50, 52 and TGF-β
52, 55, suggestive of myofibroblast differentiation.
Human cardiac fibroblasts have also been shown to differentiate to myofibroblasts after exposure
to prolonged 1% O2 for 4-8 days51. However, many details of the human cardiac fibroblast
response to hypoxia remain unknown, including their response on acute time scales and
mechanisms of their hypoxia response. For example, all cells sense and respond to hypoxia
primarily through a family of transcription factors known as hypoxia inducible factors (HIFs), which
activate gene expression programs associated with survival in low oxygen. HIFs are likely
important regulators of human cardiac fibroblast and myofibroblast phenotypes, but these

34
relationships have been incompletely characterized. One recent study showed that HIF-1 inhibits
proliferation of mouse cardiac fibroblasts post-infarction and mitigates fibrosis181, highlighting the
importance of HIF-1 in post-MI remodeling and motivating more investigation into these pathways
in a human context.
To establish a better understanding of human cardiac fibroblast phenotypes induced by
acute hypoxia, we exposed human adult cardiac fibroblasts in vitro to hypoxia (0% O2) for 4 or
24h. We then characterized phenotypes using targeted (immunostaining, RT-qPCR) and nontargeted (proteomics) approaches. In response to 4h of hypoxia, HIF-1, but not HIF-2,
accumulated in the nuclei, followed by increased expression of known downstream targets, such
as VEGFA and LOX, by 24h. Hypoxia also reduced proliferation and many pathways associated
with global protein synthesis, such as translation initiation, elongation, and termination. TGF-β1,
a positive control for activation into the myofibroblast phenotype, similarly reduced proliferation of
fibroblasts but, in contrast, upregulated various fibrotic genes, proteins, and pathways, including
α-SMA expression and collagen biosynthesis. These data suggest that acute hypoxia represses
protein synthesis, including expression of fibrotic proteins, in human adult cardiac fibroblasts.
2.2. Materials and Methods
2.2.1. Fabrication of PDMS Substrates
Polydimethylsiloxane (PDMS) was prepared by combining the base and curing agent of
the Sylgard 184 Silicone Elastomer Kit (Dow, Midland, MI, USA) in a 10:1 mass ratio and then
mixing with a planetary centrifugal mixer (AR-100, Thinky, Japan). PDMS was then spin-coated
onto 22 x 22 mm glass coverslips (Electron Microscopy Sciences, Hatfield, PA, USA) using a
G3P-8 spin-coater (Specialty Coating Systems, Indianapolis, IN, USA) with spin settings to
achieve 10-µm thickness. This recipe has been previously characterized to have an elastic
modulus of 2.7 MPa118. PDMS-coated coverslips were cured in a 65 °C oven for 4 hours, UVO-

35
treated for 8 minutes to promote protein adsorption without denaturation, and then uniformly
coated with 150 µL of human fibronectin (50 µg/mL, Corning, Corning, NY, USA) for 1 hour.
Residual fibronectin was removed with compressed air and coverslips were immediately seeded
with cells.
2.2.2. Cell Culture
Primary human adult cardiac fibroblasts (Lot 3131, Cell Applications Inc., San Diego, CA,
USA) are from the healthy ventricle of a 32-year-old Caucasian female. Cells were thawed into a
75 cm2 cell culture flask and cultured in low glucose DMEM (1g/L glucose) supplemented with
10% v/v fetal bovine serum (Gibco, Waltham, MA, USA) and 1% v/v penicillin-streptomycin. Cells
were passaged in a 1:10 ratio at 70-80% confluence into new 75 cm2 cell culture flasks, by
incubating cells with trypsin-EDTA for 5 minutes at 37 °C, neutralizing with 10% serum, and
centrifuging the solution at 400g for 5 minutes at 4 °C to remove the trypsin-containing
supernatant. For experiments, cells at passages four through eight were seeded onto PDMS
coverslips pre-coated with fibronectin at a density of 22,500 cells/cm2
. The following day, cells
were serum arrested for 48 hours with low glucose DMEM (1 g/L glucose) supplemented with
0.1% v/v fetal bovine serum and 1% v/v penicillin-streptomycin, and then exposed to hypoxia or
transforming growth factor-β1 (TGF-β1) treatment.
After 48h of serum arresting, untreated samples were used immediately for
immunostaining or RNA and protein isolation. For hypoxia treatment, coverslips were placed in a
hypoxia chamber (STEMCELL Technologies, Vancouver, BC, Canada) that was filled with 0%
O2, 5% CO2, bal N2 and incubated at 37 °C for 4 or 24h. Cells were removed from the chamber
at the designated timepoints and immediately used for immunostaining or RNA and protein
isolation. For TGF-ꞵ1 treatment, serum-arrested growth medium was supplemented with TGF-ꞵ1

36
(R&D Systems, Minneapolis, MN, USA) at 10 ng/mL for 48h prior to immunostaining or RNA and
protein isolation.
2.2.3. Immunostaining
Cells were fixed with 4% paraformaldehyde for 10 minutes then permeabilized with 0.5%
Triton X-100 in PBS for 10 minutes. Cells were then incubated with a primary staining solution
overnight on a rocker at 4°C containing either monoclonal mouse HIF-1⍺ antibody (1:200, Novus
Biologicals, Littleton, CO, USA), monoclonal rabbit Ki-67 (1:200, Invitrogen, Carlsbad, CA, USA),
monoclonal rabbit HIF-2⍺ (1:200, Novus Biologicals) or monoclonal mouse ⍺-SMA (1:200, Sigma
Aldrich, St. Louis, MO, USA). Cells were rinsed with PBS and then incubated at room temperature
for 1.5h with either Alexa Fluor 547 goat anti-mouse secondary antibody (1:200, Invitrogen,
Carlsbad, CA, USA), Alexa Fluor 547 goat anti-rabbit 488 (1:200, Invitrogen, Carlsbad, CA, USA),
DAPI (1:200, Life Technologies, Carlsbad, CA, USA), or Alexa Fluor 488 Phalloidin (1:200, Life
Technologies, Carlsbad, CA, USA). Cells were mounted onto glass slides using Prolong Gold
Anti-Fade (Invitrogen, Carlsbad, CA, USA).
2.2.4. Microscopy and Image Analysis
Fluorescent images were captured using a 60x oil objective on a Nikon Eclipse Ti-S
inverted fluorescent microscope at five distinct and random locations along the coverslips. Images
were analyzed using ImageJ. To analyze HIF-1α and HIF-2α nuclear to cytosolic intensity ratios,
a nuclear mask or its inverse were used to create images of nuclear or cytosolic spaces.
Thresholding followed by a measurement of mean pixel intensity on the thresholded region were
divided to compute the final ratios. To compute percent Ki-67 positive cells, the number of Ki-67+
cells and the total number of nuclei were counted using the ImageJ analyze particles function. To
compute percent α-SMA positive cells, the number of nuclei that overlapped with α-SMA fibers

37
were manually counted, and the total number of nuclei were counted using the ImageJ analyze
particles function.
2.2.5. RNA and Protein Isolation
TRIzol-chloroform phase separation was used for the dual extraction of mRNA and protein
from the aqueous and organic phases, respectively. mRNA transcripts were isolated using a
Qiagen miRNA Mini Kit (Bio-Rad, Hercules, CA, USA). Protein pellets were precipitated from the
organic phase using isopropanol followed by centrifugation. The pellet was washed with methanol
to remove residual contaminants and then resuspended in a buffer containing 8 M Urea/5% SDS.
2.2.6. Quantitative RT-PCR
RNA for quantitative RT-PCR (RT-qPCR) was analyzed for concentrations and 260/280
ratios using a Nanodrop (Thermo Fisher Scientific). To ensure sample purity, mRNA
concentrations above 50 ng/uL and 260/280 ratios above 1.8 were used for cDNA synthesis.
cDNA was synthesized using the iScript Reverse Transcription Supermix for RT-qPCR kit (BioRad Laboratories, Hercules, CA, USA). RT-qPCR was performed using the SsoAdvanced
Universal SYBR Green Supermix (Bio-Rad Laboratories, Hercules, CA, USA), cDNA template,
and primers for ACTA2, COL1A1, TGFB1, VIM, GAPDH and VEGFA (Bio-Rad Laboratories,
Hercules, CA, USA). The CFX384 Touch Real-Time PCR Detection System (Bio-Rad
Laboratories, Hercules, CA, USA) was used to obtain cycle threshold (Ct) datasets for each target
gene within each sample. Gene expression was normalized relative to the housekeeping gene
GAPDH. Average expression values were computed using the delta delta Ct method182
.

38
2.2.7. Proteomic Analysis
Protein concentration was determined by bicinchoninic acid (BCA) assay and 50 μg of
protein per sample was aliquoted for further processing. Protein aliquots were processed,
digested with trypsin, and cleaned using the ProtiFi S-Trap system (ProtiFi, Fairport, NY, USA).
Peptides were dried following elution from the S-Trap columns, and dried peptides were
resuspended at a concentration of 1 µg/µL for injection onto mass spectrometry (MS) system.
The samples were analyzed by 60 minute data independent acquisition (DIA) on an
Orbitrap Lumos Tribrid Mass Spectrometer (Thermo Fisher Scientific). DIA scans spanned 400-
1000 m/z with 15 m/z nonoverlapping windows and were acquired at 15,000 resolution by
accumulating ions for a maximum injection time of 30 ms to an automatic gain control (AGC)
target of 200,000 and fragmenting the ion packet by HCD at 30% collision energy. After every 40
DIA scans, a precursor (MS1) scan was acquired at 120,000 resolution with AGC target set to
600,000 and a maximum injection time of 50 ms. Electrospray ionization was carried out using a
Newomics 8 nozzle (10 um internal diameter) M3 emitter at a voltage of 3900 volts with the sheath
gas set to 20.
Liquid chromatography separation was carried out using a trap and elute configuration on
an Ultimate 3000 through a CapLC 50 cm micropillar array analytical column and a Phenomenex
trapping column. Samples were loaded using a full loop injection with a 20 uL loop at 20 uL/min
onto the trapping column. The reversed phase gradient (mobile phase A: 0.1% formic acid in
water, mobile phase B: 0.1% formic acid in acetonitrile) was delivered at 9.5 uL/min as follows:
start at 1% B and hold for 3 minutes, ramp to 4% B at 3.1 minutes, linear gradient to 15% B at 30
minutes, linear gradient to 28% at 52 minutes, ramp to 45% B at 59 minutes. After each run a
separate 10 minute equilibration method was used to wash the analytical and trapping columns
with 95% B and equilibrate the system to 1% B. Raw files were analyzed by library free search

39
using the DIANN workflow183, with RT-dependent total intensity normalization, maxLFQ proteinlevel intensity estimation, and match between runs.
Normalized area under the curve intensities were imported into the Perseus software
platform184 for statistical analysis. A t-test was performed on TGF-ꞵ1 versus untreated controls
and 24h hypoxia versus untreated controls. Proteins that met log ratio and q-value cutoffs of 0.5
and 0.05 were used in pathway analysis using Ingenuity Pathway Analysis software (QIAGEN,
Hilden, Germany). Pathways were then filtered with -log10(p-value) and z-score cutoffs of 1.3 and
1.5, respectively. To identify overlapping pathways in 24h- and TGF-ꞵ1-treated cells, shared
pathways that met -log10(p-value) cutoffs of 1.3 in both conditions were plotted using Python
scripts with pandas and seaborn packages and visualized on Jupyter notebook.
2.3. Results
2.3.1. Hypoxia Drives Nuclear Translocation of HIF-1α within Hours
The cellular response to hypoxia is mediated by hypoxia-inducible factors (HIF), a family
of transcription factors that translocate to the nucleus to initiate the transcription of genes that
promote survival in low oxygen, such as angiogenesis and glycolysis185. To characterize the
timeline of HIF activation in primary human adult cardiac fibroblasts, we cultured cells in hypoxia
(0% oxygen) for 4h or 24h, with ambient 21% O2 or TGF-β1 treatment (10 ng/µL) as untreated
and fibrotic control conditions, respectively. We also deprived all cells of serum and glucose to
minimize basal levels of fibroblast differentiation, similar to previous studies118. We then used
immunofluorescence to quantify the nuclear and cytosolic localization of HIF-1α and HIF-2α, the
alpha subunits of two key HIF isoforms.
Hypoxia resulted in increased HIF-1α accumulation in the nuclei by 4h, which subsided
after 24h (Figure 2-1A). HIF-2α, a closely related HIF isoform that is more commonly associated

40
with the chronic response to hypoxia186, was already present in the nuclei of untreated cells. This
may be because HIF-2 is less efficiently hydroxylated and degraded in normoxia186 or possibly
more sensitive to glucose deprivation. HIF-2 nuclear localization was also unchanged by 4h or
24h of hypoxia (Figure 2-1B). Thus, HIF-1 but not HIF-2 was responsive to acute hypoxia in
human adult cardiac fibroblasts. TGF-β1 treatment did not affect HIF-1 or HIF-2 activation
compared to no treatment (Figure 2-1A and B), as expected.
Figure 2-1. HIF-1 and HIF-2 localization in human cardiac fibroblasts treated with hypoxia or TGFβ1. Immunofluorescence of (A) HIF-1 (red) and DAPI (blue) and quantification of HIF-1
nuclear/cytosolic intensity. (B) HIF-2 (red) and DAPI (blue) and quantification of HIF-2
nuclear/cytosolic intensity. Brightness and contrast are adjusted on all images. Scale bars = 100
µm. NT = no treatment. Each data point is the average of 5 images per coverslip across n=4
coverslips from 3 independent trials. ****p<0.0001 according to ordinary one-way ANOVA with
Dunnett’s multiple comparisons with NT as the control.
2.3.2. Hypoxia Inhibits Cardiac Fibroblast Proliferation
We next asked if cardiac fibroblast activation and/or proliferation are affected by acute
hypoxia. To answer this, we stained cells for α-SMA, a marker of myofibroblasts, or Ki-67, a
marker of proliferation. We found that the percentage of α-SMA-positive cells was unchanged by

41
4h and 24h of hypoxia (Figure 2-2A). In contrast, TGF-β1 treatment caused an increase in α-SMA
expression, as expected (Figure 2-2A). Proliferation was inhibited by both TGF-β1 and hypoxia
(Figure 2-2B) compared to no treatment. Hypoxia187 and TGF-β1188 have previously been shown
to cause cell cycle arrest in many other cell types, consistent with our results.
Figure 2-2. Differentiation and proliferation of human cardiac fibroblasts treated with hypoxia or
TGF-β1. Immunofluorescence of (A) α-SMA (red), actin (green), DAPI (blue) and quantification of
α-SMA+ cells. Immunofluorescence of (B) Ki-67 (green) and DAPI (blue) and quantification of
Ki67+ nuclei. Brightness and contrast are adjusted on all images. Scale bars = 100 µm. NT = no
treatment. Each data point is the average of 5 images per coverslip across n=4 coverslips from 3
independent trials. *p<0.05, **p<0.01, according to ordinary one-way ANOVA with Dunnett’s
multiple comparisons with NT as the control.
2.3.3. Hypoxia Drives Expression of the HIF Target Gene VEGFA
We next measured the expression of several hallmark genes of hypoxia and fibrosis using
RT-qPCR. The gene that encodes for vimentin (VIM), a non-specific marker of both fibroblasts
and myofibroblasts189, 190, was not significantly different between conditions (Figure 2-3A).
Hypoxia upregulated vascular endothelial growth factor (VEGFA), a known target of HIF-1
191
, by
24h (Figure 2-3B). Conversely, fibrotic target genes that encode α-SMA (ACTA2), type 1 collagen

42
(COL1A1) and TGF-β1 (TGFB1) were all upregulated by TGF-β1 but not hypoxia (Figure 2-3CE). Because expression of these genes is consistent with myofibroblast differentiation, these
results further suggest that TGF-β1, but not acute hypoxia induces myofibroblast differentiation.
Figure 2-3. Expression of hypoxic and fibrotic target genes in human cardiac fibroblasts treated
with hypoxia or TGF-β1. Quantitative RT-PCR of genes encoding (A) vimentin, (B) vascular
endothelial growth factor, (C) alpha-smooth muscle actin, (D) collagen type I, and (E) transforming
growth factor β1 after hypoxia (0% O2) or TGF-β1 treatment. NT = no treatment. Each data point
reflects two pooled coverslips per trial from n=4 independent trials. *p<0.05, ***p<0.001,
****p<0.0001 according to ordinary one-way ANOVA with Dunnett’s multiple comparisons with NT
as the control.
2.3.4. Hypoxia Downregulates Collagen Synthesis
Finally, we conducted a proteomic analysis in order to more broadly identify changes in
fibroblast phenotype induced by hypoxia or TGF-β1 treatment. We quantified 5079 proteins and
found that TGF-β1 and hypoxia significantly changed the expression of 479 and 827 proteins,
respectively (Figure 2-4A). Of these proteins, 155 were shared by both conditions, 142 of which
changed in the same direction. Consistent with our RT-qPCR results, TGF-β1 treatment increased
expression of TGF-β1 (TGFB1) and collagens (CO1A1 and CO1A2) (Figure 2-4B-D). In contrast,

43
hypoxia decreased expression of CO1A1 and CO1A2 but increased expression of lysyl oxidase
(LOX), a known target of HIF-1 in cancer cells which cross-links collagen fibers192. We also
observed increased expression of YAP1, a transcriptional co-activator, only in fibroblasts treated
with TGF-β1 (Figure 2-4E). YAP is a transcription factor activated in fibroblasts after myocardial
infarction193, 194 and in response to stiffness120. Together, these changes in protein expression
further demonstrate that TGF-β1, but not acute hypoxia, induced expression of fibrotic targets.
Figure 2-4. Proteomic analysis of human cardiac fibroblasts treated with hypoxia or TGF-β1. (A)
Venn diagram comparing significant proteins after hypoxia or TGF-β1 treatment. Normalized
intensities for (B) TGFB1, (C) CO1A1, (D) CO1A2, (E) YAP1, and (D) LOX. NT = no treatment.
Each data point reflects one of n=5 or 6 coverslips from 4 independent trials. *p<0.05, ***p<0.001,
****p<0.0001 according to ordinary one-way ANOVA with Dunnett’s multiple comparisons with NT
as the control.
2.3.5. Hypoxia Globally Represses Protein Synthesis in Cardiac Fibroblasts
To characterize the functional implications of proteomic differences, we performed
pathway analysis on differentially expressed proteins that passed p-value and fold change cutoffs
of 0.05 and 1.5, respectively. As shown in Figure 2-5A, hypoxia induced a stress response that
reduced many pathways related to global protein synthesis. Downregulated pathways include

44
decreased Eukaryotic Initiation Factor 2 (EIF2) Signaling, Eukaryotic Translation Initiation,
Elongation, and Termination, and rRNA Processing, which are all canonical pathways involved in
eukaryotic protein synthesis195. In contrast, TGF-β1 treatment upregulated many pathways. The
top 20 enriched pathways, sorted by p-value, include Ephrin Receptor Signaling, which has
previously been implicated in myofibroblast activation196, and several pathways associated with
cell adhesion and cytoskeletal organization (Actin Cytoskeleton Signaling, Integrin Signaling,
Signaling by Rho Family GTPases) and cell junction signaling (Cell Junction Organization,
Remodeling of Epithelial Adherens Junctions) (Figure 2-5B).
Next, we plotted shared pathways with a dot plot, in which dot size denotes significance
and color denotes z-score (Figure 2-5C). Shared pathways were included if both comparisons
passed a p-value threshold of 0.05. Of the shared pathways, TGF-β1-treated cells upregulate
Pulmonary Fibrosis, while hypoxia-treated cells downregulate this pathway (Figure 2-5C).
Hypoxia-treated cells also showed an increase in Degradation of the Extracellular Matrix, which
may facilitate cell migration (Figure 2-5C).

45
Figure 2-5. Pathway analysis of human cardiac fibroblast proteomes after treatment with 24h
hypoxia or TGF-β1. Top 20 pathways, sorted by p-value, in cells treated with (A) hypoxia or (B)
TGF-β1, after applying p-value and absolute z-score cutoffs of 0.05 and 1.5, respectively. (C) Dot
plot reflecting shared pathways that pass p-value cutoffs of 0.05 in both conditions, where size
represents significance and color denotes z-score.

46
We next sorted our proteomic data to identify proteins based on fold-change or statistical
significance. Among the top proteins identified in hypoxic cells is angiopoietin-like 4 (ANGPTL4),
ferritin heavy chain 1 (FTH1), and N-myc downstream regulated gene 1 (NDRG1) (Figure 2-6AC), ANGPTL4 attenuates angiotensin-II induced proliferation, migration, α-SMA expression, and
collagen production in rat atrial fibroblasts197 and may play a similar role in suppressing fibrosis
in hypoxic fibroblasts. Interestingly, ANGPTL4 was not detected in TGF-β1-treated cells. FTH1 is
the heavy chain unit of ferritin, a major iron storage protein, and has been shown to inhibit
proliferation in other cell types198
. Interestingly, it was also upregulated in TGF-β1-treated cells.
Hypoxia, but not TGF-β1, also caused an increase NDGR1, a target of HIF-1 that is also known
to inhibit proliferation199. Thus, proliferation may be inhibited in cells exposed to hypoxia or TGFβ1 via distinct mechanisms.
Among the top proteins identified in TGF-β1-treated cells was connective tissue growth
factor (CCN2), which drives myofibroblast differentiation in response to TGF-β
200, and palladin
(PALLD), an actin associated protein involved in smooth muscle cell differentiation and
migration201 (Figure 2-6D-E). We also saw upregulation of acetyl co-A binding protein (ACBP)
(Figure 2-6F), which regulates autophagy and is correlated with cardiovascular risk factors202
.
Interestingly, we also observed upregulation of proteins that are related to the response to hypoxia
or oxidative stress, including LanC like glutathione S-transferase 2 (LANCL2), pyruvate kinase
L/R (PKLR), and prion protein (PRIO) (Figure 2-6G-I). PKLR is involved in glycolysis in liver and
red blood cells. LANCL2 is a receptor for abscisic acid and has been shown to be upregulated to
protect cardiac myocytes from hypoxic injury via nitric oxide production203. Lastly, PRIO can
reduce cardiac oxidative stress204 and is correlated with cellular proliferation in a mouse model of
transverse aortic constriction205. Although TGF-β1-treated cells were cultured in ambient oxygen,
oxidative stress206 and TGF-β1207 have been shown to promote a pseudohypoxic state regardless

47
of oxygen level in lung fibroblasts, which may be similar to some of the phenotypes we observed
in TGF-β1-treated cardiac fibroblasts.
Figure 2-6. Expression of proteins with highest significance or fold change in the proteomic
analysis of human cardiac fibroblasts treated with hypoxia or TGF-β1. Normalized intensites of
(A) Angiopoietin-like 4, (B) ferritin heavy chain 1, (C) N-myc downstream regulated gene 1, (D)
connective tissue growth factor, (E) palladin, (F) acetyl co-A binding protein, (G) pyruvate kinase
L/R, (H) LanC like glutathione S transferase 2, and (I) prion protein. NT = no treatment. Each data
point is one of n=5 or 6 coverslips from 4 independent trials. *p<0.05, ***p<0.001, ****P<0.0001
according to ordinary one-way ANOVA with Dunnett’s multiple comparisons with NT as the
control.
2.4. Discussion
Myocardial infarctions trigger a complex wound healing process that commonly results in
pathological fibrosis by mechanisms that are incompletely understood. Because the post-infarct

48
microenvironment is very complex, in vitro models are well-suited for teasing apart the impact of
distinct factors on individual cell populations. In this study, our goal was to evaluate the acute
effects of hypoxia and TGF-β1 on human cardiac fibroblast phenotypes, including the activity of
HIF transcription factors and gene and protein expression.
We found that human cardiac fibroblasts responded to hypoxia with increased HIF-1
localization to nuclei within hours. This was followed by increased VEGFA gene expression and
LOX protein levels by 24h, which are both known to be downstream targets of HIF-1
191, 192. We
also observed decreased proliferation, which has been shown to be suppressed by HIF-1 in
cardiac fibroblasts181 and by NDRG1, a target of HIF199, both of which were upregulated in hypoxic
cells. We also observed a global repression of protein synthesis, including decreased signaling
by EIF2. EIF2, or eukaryotic initiation factor 2, is involved in global initiation of translation195 and
is phosphorylated to inhibit protein synthesis in response to environmental stressors208. This was
accompanied by no change or decreases in expression of genes and proteins related to
myofibroblast differentiation and matrix synthesis in hypoxic cells, including decreased expression
of collagen protein, no change in ACTA2 or TGFB1 expression, and a decrease in the pathways
for collagen biosynthesis and pulmonary fibrosis. Pulmonary and cardiac fibrosis are known to
share many cellular and molecular mechanisms209. Among the highest fold-change proteins in
hypoxic cells, we also saw increased ANGPTL4, which is a known target of HIF-1 in melanoma
cells210 and has been shown to inhibit collagen synthesis and α-SMA expression in fibroblasts197
,
which may also suggest suppression of myofibroblast differentiation at the protein-level. While
other studies have shown that hypoxia induces differentiation of fibroblasts into myofibroblasts50-
52, 55 these have not looked at human fibroblast activation in response to short-term hypoxia
treatment. It is possible that hypoxic activation of fibroblasts takes longer than 24h, which was our
endpoint. In vivo, fibroblast activation occurs after 3-7 days211
. Hypoxic suppression of protein
synthesis, which was observed in this study, may be an initial step towards differentiation of
fibroblasts to myofibroblasts. Translation has been shown to be reprogrammed during cellular

49
differentiation in other cell types. For example, CD8 effector T cells increase translation during
proliferation but repress translation during differentiation212. Translational repression is also
induced by hypoxia in cancer213, 214. Differences between in vitro studies may also be caused by
cell culture conditions. For example, we deprived fibroblasts of glucose and serum to minimize
basal levels of fibroblast differentiation100. Because cardiac fibroblasts are a heterogeneous
population of cells with phenotypic plasticity, factors such as cell source, serum levels, oxygen
concentration, and hypoxia duration likely all affect fibroblast phenotype and are often not
consistent across studies. Together, our data suggest that hypoxia inhibited proliferation and
protein synthesis in human cardiac fibroblasts on acute time scales. In response to TGF-β1, a
positive control for fibroblast activation, cardiac fibroblasts demonstrated hallmark signs of
myofibroblast differentiation, including upregulation of TGF-β1, collagen type 1, and α-SMA, as
expected. We also interestingly observed increased protein levels of YAP-1, a key driver of
stiffness-dependent activation of fibroblasts120. This could indicate that TGF-β1 induced a change
in apparent stiffness due to matrix deposition. We also observed enrichment of pathways related
to cell junctions in TGF-β1-treated cells. Electrotonic communication through cell-cell junctions
has been observed between cardiac fibroblasts and myocytes in animal215-218 and increases after
injury216, although it is not known whether this occurs in humans219
.
Surprisingly, we also observed a few proteins and pathways in fibroblasts treated with
TGF-β1 that are related to hypoxia responses. For example, TGF-β1 treatment upregulated the
pathway for translocation of GLUT4 to the plasma membrane. TGF-β1 treatment also upregulated
PKLR, a pyruvate kinase, and LANCL2, which together with LANCL1 serves as a receptor for the
stress hormone abscisic acid and is involved in the response of cardiac myocytes to hypoxia203
.
Previous studies have similarly observed expression of hypoxia-associated genes and proteins
regardless of oxygen level and coined this phenotype “pseudohypoxia”220, 221. Pseudohypoxia is
a common feature of cancer, where cells shift towards glycolysis despite normal oxygen levels
and functioning mitochondria222. Pseudohypoxia has also been observed in lung fibrosis206, 223 and

50
cardiac aging224. For example, HIF signaling is active in fibroblasts isolated from patients with
pulmonary fibrosis206, 223 and normal lung fibroblasts augment glycolysis and HIF signaling in
response to TGF-β treatment for 6-24h207. Although we did not assess HIF-1 immunofluorescence
until after the full 48 hours of TGF-β1 treatment, future studies can explore HIF-1 immunostaining
in TGF-β1-treated cardiac fibroblasts at earlier timepoints to further explore this phenomenon.
In this study, we utilized primary human adult cardiac fibroblasts to preserve human
relevance. However, one limitation is that we used only one patient line, which may not be
representative of the broader population. We also did not assess the combined effects of TGFβ1 and hypoxia, which have been shown to synergistically increase myofibroblast markers in lung
fibroblasts207. Our simplified in vitro system also does not include other key cell types in the postinfarct microenvironment, including immune cells and cardiac myocytes. Despite these limitations,
these data delineate the unique roles of hypoxia and TGF-β1 in regulating cardiac fibroblast
phenotype in a simplified system and provide a more human-relevant characterization of acute
cardiac fibroblast responses to stimuli present due to infarction, helping pave the way for future
therapeutic interventions to mitigate pathological fibrosis.

51
Chapter 3 Hypoxic-Normoxic Crosstalk under Microfluidic Control Activates Unique
Inflammatory Pathways in Human Cardiac Fibroblasts
This Chapter is an adaptation of: Khalil Natalie N., Rexius-Hall Megan L., Gupta Divya, McCarthy
Liam, Verma Riya, Kellogg Austin C., Takamoto Kaelyn, Xu Maryann, Nejatpoor Tiana, Parker
Sarah J., and McCain Megan L. Crosstalk with Hypoxic Cardiac Fibroblasts Activates ProInflammatory Signaling in a Post-Infarct Human Myocardium on a Chip. Advanced Healthcare. In
Review.
Myocardial infarctions locally deprive myocardium of oxygenated blood and cause
immediate cardiac myocyte necrosis. Irreparable myocardium is then replaced with a scar through
a dynamic repair process that is an interplay between hypoxic cells of the infarct zone and
normoxic cells of adjacent healthy myocardium. In many cases, unresolved inflammation or
fibrosis occurs for reasons that are incompletely understood, increasing the risk of heart failure.
We hypothesized that crosstalk between hypoxic and normoxic cardiac fibroblasts regulates
mechanisms of repair after a myocardial infarction. To test this, we fabricate microfluidic devices
on 3-D printed templates for co-culturing hypoxic and normoxic cardiac cells. We found that
hypoxia drives gene expression changes in human cardiac fibroblasts that are associated with
glycolysis and a pro-fibrotic phenotype, similar to the anti-inflammatory phase of wound healing.
Co-culture with normoxic fibroblasts uniquely upregulated pro-inflammatory signaling in hypoxic
fibroblasts, including increased secretion of TNF-α. Together, these data suggest that crosstalk
between hypoxic and normoxic fibroblasts uniquely activates phenotypes that resemble the initial
pro-inflammatory phase of post-infarct wound healing.
3.1. Introduction
Myocardial infarctions acutely deprive downstream myocardium of oxygenated blood,
resulting in the necrosis of up to one billion cardiac cells within hours12, 225, 226. Injured myocardium
then undergoes a wound healing process comprising inflammatory, proliferative, and maturation
phases. During the inflammatory phase, innate immune cells are recruited to the site of injury by

52
pro-inflammatory chemokines, such as the C-C and C-X-C subfamilies227, 228, and cytokines, such
as IL-1
29, 229, 230, to remove cell and matrix debris. Immune cells, fibroblasts, and/or cardiac
myocytes may serve as sources of these cytokines and chemokines, although the relative
contributions of each cell type are not clear29. Cardiac fibroblasts29, 177 also play a pivotal role
during the proliferative and maturation phases by proliferating, differentiating into myofibroblasts,
and depositing new matrix to form a scar. Pro-inflammatory cytokines, such as IL-1, initially
repress fibrosis in the infarct microenvironment230. At the later proliferative stage, antiinflammatory mediators are secreted, such as IL-10 and TGF-β1, to promote the differentiation of
fibroblasts into myofibroblasts231. Importantly, timely resolution of both inflammation and fibrosis
are essential to minimize the extent of injury after myocardial infarction. Excessive and unresolved
fibrosis can weaken or impair synchronous contraction of the myocardium, leading to arrhythmias
or even heart failure. Pathological fibrosis post-myocardial infarction continues to be a major
clinical problem because the underlying mechanisms that drive and suppress inflammation and
fibrosis are still not fully understood, in large part due to the spatial and temporal complexities of
post-infarct remodeling232
.
One of the most striking changes to the cellular microenvironment after a myocardial
infarction is localized oxygen deprivation. The impact of hypoxia on cellular phenotypes has
conventionally been measured by exposing cells to a single, global oxygen concentration using a
hypoxia chamber. These studies have found that low oxygen drives cardiac fibroblast activation
into a myofibroblast phenotype, with increased α-SMA expression50-52, collagen production51-53
,
and migration capacity55. In general, hypoxia inhibits the proliferation of many cell types187, though
studies with cardiac fibroblasts have shown both increased50, 51, 63 and decreased proliferation in
response to hypoxia. Hypoxic fibroblasts have also been shown to secrete both pro-inflammatory
mediators, such as IL-6
63, and anti-inflammatory mediators, such as TGF-β152. The functional
impact of factors secreted by hypoxic cardiac fibroblasts has also been investigated by
transferring conditioned media to other types of cells in culture59-61, 64. However, neither hypoxia

53
chambers nor conditioned media experiments replicate the ongoing crosstalk that occurs between
hypoxic and normoxic cells in the true infarct border zone. Thus, continuous interactions and
feedback loops between hypoxic and normoxic cells cannot be detected, which may be important
regulators of the wound healing process and could be effective pathways therapeutically target
and mitigate post-infarct fibrosis.
To delineate the roles of distinct cell types in post-infarct remodeling, in vitro models that
can closely model the spatial and temporal complexities of the post-infarct oxygen landscape are
needed. Recently, microfluidic devices that enable oxygen control at the microscale have been
developed that enable the direct co-culture of hypoxic and normoxic cells70, 233, 234. Here, we
engineered microfluidic devices from 3-D printed templates to identify oxygen-dependent
crosstalk in human cardiac fibroblasts. We found that hypoxia induced several expected
responses, including HIF-1-accumulation in nuclei and upregulation of glycolysis, regardless of
co-culture with normoxic cells. Hypoxia also universally activated pathways for wound healing
and anti-inflammatory signaling, which are generally associated with the proliferative phase of
post-infarct remodeling. Interestingly, hypoxic cells in co-culture with normoxic cells uniquely
upregulated pro-inflammatory signaling pathways, such as macrophage classical activation, as
well as secretion of TNF-α. Together, our results suggest that crosstalk between normoxic and
hypoxic cardiac fibroblasts may drive unique signaling pathways related to immune cell
recruitment and activation that resemble the initial inflammatory phase post-infarction.
3.2. Materials and Methods
3.2.1. Microfluidic Device Fabrication
Microfluidic channels for gas flow were designed in Autodesk Inventor (Autodesk, Inc.,
San Francisco, CA, USA) with a width of 1000 µm wide and depth of 100 µm. Channels were
arranged into a serpentine network with a separation of 500 µm. Each serpentine network was

54
roughly 20 x 22 mm. The inverse design was 3D printed in Master Mold Photopolymer Resin
using a CADworkds3D μMicrofluidics Edition M-50 3D Printer (CADworks3D, Concord, ON,
Canada). 3D-printing enabled rapid prototyping that allowed us to sample various iterations of
device geometries before deciding on our simple channel network that maximizes cell surface
area (Figure 3-1A). After printing, resin pieces were processed with two successive baths in
isopropyl alcohol, followed by detailing with a spray gun loaded with isopropyl alcohol. Resin
pieces were then cured for 40 minutes using the CureZone UV curing unit (CADworks3D,
Concord, ON, Canada) and placed in a 65°C dry oven overnight. Resin pieces were then filled
with polydimethylsiloxane (PDMS), which was prepared by mixing the base and curing agent of
the Sylgard 184 Silicone Elastomer Kit (Dow, Midland, MI, USA) in a 10:1 mass ratio with a
planetary centrifugal mixer (AR-100, Thinky, Japan). PDMS-filled molds were degassed for 30
minutes in a vacuum chamber, flattened with a 150 mm petri dish lid, and then cured at 65°C.
PDMS replicas were then biopsy punched at the inlets and outlets for integration with tubing for
gas flow. To seal both the channels and biopsy punched outlets, PDMS membranes were plasma
bonded to PDMS replicas by exposing to plasma for 90 seconds with a plasma cleaner (Harrick,
Ithaca, NY). PDMS membranes were made by spin coating PDMS onto blank 3-inch silicon
wafers using predefined spin settings to achieve 100-µm thickness235 with a G3P-8 spin-coater
(Specialty Coating Systems, Indianapolis, IN, USA).
3.2.2. Cell Culture Chamber Fabrication
A piece to divide the microfluidic device into three cell culture compartments was designed
in Autodesk Inventor (Autodesk, Inc., San Francisco, CA, USA). The piece is 10 mm tall, 23 mm
long, and 49 mm wide with two 1 mm-wide barriers to form three compartments. Chambers were
printed in acrylonitrile butadiene styrene (ABS) or polylactic acid (PLA) on a FlashForge Creator
Max 2 Independent Dual Extruder 3D Printer (FlashForge USA, Industry, CA, USA) and replica-

55
molded with PDMS. PDMS replicas were removed from molds by soaking in acetone. A PDMS
lid to enclose chambers was fabricated by pouring 4-5mm of PDMS in an empty 150-mm petri
dish, degassing, and curing at 65°C, similar to previous work236. To enclose chambers, chambers
were painted with uncured PDMS at the open surface, placed in contact with the PDMS-filled petri
dish, and cured at 65°C. After curing, lids were cut-to-size from the remainder of the dish.
Enclosed PDMS chambers were then bonded to microfluidic devices to form one complete device.
A flowchart depicting fabrication steps is shown in Figure 3-1B.
Prior to oxygen modulation, lids were peeled off of the device for UVO-treating, cell
seeding, and media changes, then re-attached again with uncured PDMS by curing together in a
37°C cell culture incubator. PDMS devices were UVO-treated for 8 minutes and then uniformly
coated with 150 µL of human fibronectin (50 µg/mL, Corning, Corning, NY, USA) per compartment
for 1 hour. Residual fibronectin was removed with compressed air and devices were seeded with
cells.

56
Figure 3-1. (A) Digital light processing 3D-printing enables rapid prototyping of
microfluidic networks, ranging from a 22 x 22 mm coverslip-size device that fits neatly into a sixwell plate format, a 48 mm-long device that maintains cell culture area but maximizes surface
area of the interface region, and an early prototype of a multiplexed device in which the inlets
branch on-chip to feed two adjacent channel networks (B) Device fabrication flow chart.
Fabrication entails sealing microfluidic devices with membranes, attaching chambers to lids, and
then bonding devices to chambers.

57
3.2.3. Oxygen Validation
Surface oxygen levels were validated using an optical oxygen sensor, platinum(II)
octacethylporphyrin ketone (PtOEPK), similar to previous work70. Briefly, 0.5 mg/mL of PtOEPK
was added to a 35% w/w toluene/polystyrene mixture and deposited onto the surface of the
device. The cell culture chamber was filled with 4 mL PBS and then bonded to a lid by painting
with uncured PDMS. Fluorescent images at 32 successive locations along the device were
captured using a 20x air objective on a Nikon Eclipse TI-S inverted fluorescence microscope.
Mean pixel intensity was measured over the entire image after applying a gaussian blur using
automated scripts implemented in ImageJ. Pixel intensity was used to calculate oxygen
concentration by solving the Stern Volmer relationship in MATLAB, which was calibrated in: 1)
21% O2 and 2) 0% O2 (5% CO2, balanced N2), as previously described235
.
3.2.4. Cell Culture
Primary human adult cardiac fibroblasts (Lot 3131, Cell Applications Inc., San Diego, CA,
USA) were purchased from Cell Applications and are from the normal ventricle of a 32-year-old
Caucasian female donor. Cells were thawed into a 75 cm2 cell culture flask and cultured in low
glucose DMEM (1g/L glucose) supplemented with 10% v/v fetal bovine serum (Gibco, Waltham,
MA, USA) and 1% v/v penicillin-streptomycin. Cells were passaged in a 1:10 ratio at 70-80%
confluence into new 75 cm2 cell culture flasks, by incubating cells with 1x trypsin-EDTA for 5
minutes at 37 °C, neutralizing with cell culture medium containing 10% serum, and centrifuging
the solution at 400g for 5 minutes at 4°C to remove the supernatant. For all experiments, cells at
passages four through eight were seeded onto fibronectin-coated devices at a density of 22,500
cells/cm2
.

58
3.2.5. Microfluidic Oxygen Modulation
Prior to oxygen modulation, cells were serum arrested for 48 hours with low glucose
DMEM (1 g/L glucose) supplemented with 0.1% v/v fetal bovine serum and 1% v/v penicillinstreptomycin. After 48h of serum-arresting, fresh serum-arrested media was added. PDMS lids
were then attached to devices using uncured PDMS. Each microfluidic channel network was then
connected via tubing to a compressed gas cylinder supplying 0% O2 (5% CO2, balanced N2) or
10% O2 (5% CO2, balanced N2). The flow from each gas cylinder was controlled by a regulator so
that pressure in the two microchannels was kept at 4 psi to establish stable and reproducible
oxygen levels, as described previously70
.
3.2.6. Immunostaining
Cells were fixed with 4% paraformaldehyde for 10 minutes, then permeabilized with 0.5%
Triton X-100 in PBS for 10 minutes. Cells were blocked for 1 hour at room temperature in 10%
Normal Goat Serum (Thermo Fisher Scientific, Waltham, MA, USA) then incubated with a primary
staining solution overnight on a rocker at 4°C containing either monoclonal mouse HIF-1⍺
antibody (1:200, NB100-105, Novus Biologicals, Littleton, CO, USA), monoclonal rabbit Ki-67
(1:200, MA5-14520, Thermo Fisher Scientific), monoclonal rabbit HIF-2⍺ (1:200, NB100-122,
Novus Biologicals) or monoclonal mouse ⍺-SMA (1:200, A5228, Millipore Sigma, Darmstadt,
Germany) in 10% goat serum. Cells were rinsed with PBS and then incubated at room
temperature for 1.5h with either Alexa Fluor 546 goat anti-mouse secondary antibody (1:200, A11003, Thermo Fisher Scientific), Alexa Fluor 546 goat anti-rabbit 488 (A-11008, 1:200, Thermo
Fisher Scientific), DAPI (1:200, D1306, Life Technologies, Carlsbad, CA, USA), or Alexa Fluor
488 Phalloidin (1:200, A12379, Life Technologies) in 10% goat serum. Cells were mounted onto
glass slides using 100 µL of Prolong Gold Anti-Fade (Life Technologies).

59
3.2.7. Microscopy and Image Analysis
Fluorescent images were captured using a 60x oil objective on a Nikon Eclipse Ti-S
inverted fluorescent microscope. Images were taken at five distinct and random locations along
each device region from three separate trials. All image quantifications were conducted using
ImageJ macros. To analyze α-SMA-positive pixels, a gaussian blur was used to smooth the
image, a triangle threshold was used to select the positive pixels, and a histogram was used to
count the total number of α-SMA and actin pixels, similar to previous work118. To analyze HIF-1α
nuclear to cytosolic intensity ratios, image subtraction using HIF and a nuclear mask or its inverse
was used to create images of nuclear or cytosolic spaces. The final ratios were computed by
dividing the mean pixel intensity of the thresholded nuclear region divided by the mean pixel
intensity of the thresholded cytosolic region, similar to previous work70. To compute percent Ki67-positive cells, the number of nuclei were counted using the Analyze Particles function and the
number of Ki-67+ nuclei were manually counted. Data were tested for normality using the ShapiroWilk test and then analyzed using ordinary one-way ANOVA with Tukey’s multiple comparisons
test or the Kruskal-Wallis test using GraphPad Prism software (GraphPad, La Jolla, CA, USA).
3.2.8. RNA and Cell Media Isolation
Immediately after cell culture on the device for 4 or 24h, media was transferred to
microcentrifuge tubes and stored at -80°C until further analysis. Devices were immediately cut in
half with a razor blade to separate hypoxic and normoxic co-culture regions, and all four regions
were incubated in TRIzol. TRIzol-chloroform phase separation was used to extract mRNA which
was purified using a Qiagen miRNA Mini Kit (Bio-Rad, Hercules, CA, USA).

60
3.2.9. RNA-Sequencing and Data Analysis
Each replicate used for RNA-Sequencing was pooled from three samples across six
independent trials. Total RNA samples for each replicate were analyzed for RNA integrity on the
2100 Bioanalyzer using the Agilent RNA 6000 Nano Kit (Agilent Technologies, Santa Clara, CA)
and quantified using the Qubit RNA HS Assay Kit (ThermoFisher Scientific, Waltham, MA). Up to
one μg of total RNA was purified for mRNA using the NEBNext® Poly(A) mRNA Magnetic
Isolation Module (New England Biolab Inc, Ipswich, MA). Stranded RNA-Seq library construction
was performed using the xGen Broad-Range RNA Library Prep Kit (Integrated DNA
Technologies, Coralville, IA). Library concentration was measured with a Qubit
fluorometer (ThermoFisher Scientific) and library size was evaluated on a 4200 TapeStation
(Agilent Technologies). Multiplexed libraries were sequenced on a NovaSeq 6000 (Illumina, San
Diego, CA) using 75bp single-end sequencing. On average, approximately 30 million reads were
generated from each sample.
Similar to previous work236, sequences were analyzed using Partek Flow Genomic
Analysis Software (Partek Inc., St. Louis, MO, USA), where they were first trimmed using a Phred
score of 20 and min read length 25. Then STAR (2.7.8a) was used to align reads to the hg38
genome. Aligned reads were quantified using encode 38 (V2), and counts were normalized by
median ratio normalization. For hypoxia versus normoxia and hypoxic co-culture versus normoxic
co-culture comparisons, sequences were analyzed for differential gene expression using DESeq2
method with p-value, fold change, and false discovery rate cutoffs of 0.05, 1.5, and 0.05,
respectively. For hypoxic co-culture versus hypoxia and normoxic co-culture versus normoxia, we
removed the false discovery rate cutoff but strengthened the p-value threshold to 0.01.
Differentially expressed genes that met these cutoffs were used for pathway analysis in Ingenuity
Pathway Analysis on the fold change and p-values (QIAGEN, Hilden, Germany). Raw and
processed data are available on Gene Expression Omnibus (Accession No. GSE254962).

61
3.2.10. MSD Assay
An enzyme immunoassay of cell media was performed using an MSD QuickPlex SQ
120MM Imager (Meso Scale Diagnostics, LLC., Rockville, MD, USA) according to the MSD
Human V-Plex Pro-Inflammatory 10-Plex Assay Kit (Meso Scale Diagnostics, LLC., Rockville,
MD, USA). Briefly, samples, calibrators, and controls were prepared according to kit instructions
and added to duplicate wells of a plate. Each sample was a pooling of three replicates from a total
of six independent trials. Plates were incubated with detection antibody solution, after which they
were washed and read.
3.3. Results
3.3.1. Microfluidic Devices Enable Co-Culture of Hypoxic and Normoxic Cells
In vivo oxygen probes have shown that oxygen levels at a myocardial infarct and remote
zone are 0% and 10%, respectively35. Thus, we designed a microfluidic cell culture device with a
central compartment for culturing hypoxic cells adjacent to normoxic cells, similar to previous
approaches70, 233, 234. Within this compartment, oxygen levels are spatially controlled by culturing
cells on a thin membrane of polydimethylsiloxane (PDMS) bonded on top of two adjacent
microfluidic channel networks. Gases with defined levels of oxygen are flown through the
channels, where they then diffuse across the PDMS membrane and dictate the level of oxygen
experienced by cells cultured on the surface of the membrane. Thus, hypoxic cells can be cultured
adjacent to normoxic cells by flowing 0% oxygen through one channel and 10% oxygen through
the other (Fig 3-2A). We also designed two compartments that flank the central compartment but
share only one microfluidic gas channel to expose cells to uniform 0% oxygen or uniform 10%
oxygen for direct comparison to cells in co-culture.

62
To fabricate our device, we first designed microfluidic gas supply channels as two adjacent
serpentine networks to both maximize surface area for gas exchange and provide structural
support for the cell culture membrane. Conventionally, templates for microfluidic channels, such
as these, are fabricated using photolithography, which is time- and resource-intensive. To
accelerate the process, we used a benchtop digital light processing 3D-printer to more rapidly
fabricate microfluidic channel templates from photocurable resin. We then replica molded 3Dprinted templates in PDMS and bonded a thin, 100-µm PDMS membrane on top of the channel
networks (Figure 3-2B-C). To separate the device into three distinct culture compartments, we
used a 3D-extrusion printer to print an inverse design of compartment barriers. We then replica
molded these templates in PDMS and soaked them in acetone for easier separation of PDMS
from the 3D-printed template. A complete flow chart is shown in Figure 3-1B. Finally, we bonded
the PDMS compartment barriers and lids on top of the PDMS membrane to form enclosed
chambers that contain cells and media. The complete assembly of the device with three
independent regions – the hypoxic control, hypoxic and normoxic co-culture, and normoxic control
regions – which correspond to oxygen landscapes in the infarct zone, border zone, and remote
zones post-infarction, is shown in Figure 3-2C.
To validate that our device could spatially regulate oxygen levels, we coated devices with
an optical oxygen sensor, platinum(II) octaethylporphyrinketone (PtOEPK), which is fluorescent
in the absence of oxygen and quenched in the presence of oxygen. We then continuously
perfused one channel with 0% oxygen and the other with 10% oxygen and measured
fluorescence intensity across the device. As shown in Figure 3-2D, fluorescence was high above
the 0% oxygen channel and low above the 10% oxygen channel, as expected. We then used
fluorescence intensities to calculate oxygen concentrations by solving the Stern-Volmer
relationship, as previously described70, 233, 237. As shown in Figure 3-2D, half the device
experienced 0% oxygen and the other half experienced 10% oxygen with a relatively steep
transition, matching our microfluidic channel design.

63
Figure 3-2. Microfluidic Devices to Co-Culture Hypoxic and Normoxic Cells. (A) Microfluidic
devices contain two channels for gas flow that mimic infarct and remote zones in post-infarct
myocardium. Gases diffuse across a 100 µm membrane to reach cells cultured on its surface. (B)
Channel networks are DLP 3D-printed and replica molded in PDMS. (C) Complete assembly
involves bonding to a PDMS membrane, chamber and lid. Chambers multiplex the device into
hypoxia, normoxia, and co-culture regions. (D) An optical oxygen sensor, PtOEPK (red),
demonstrates adjacent regions of 0 and 10% oxygen along the device from n=3 independent
trials. Schematic in A is Created in Biorender.com.
3.3.2. HIF Activation is Restricted to Hypoxic Cardiac Fibroblasts Regardless of Co-Culture
Next, we used our device to determine if cardiac fibroblast phenotypes are affected by
hypoxia and/or crosstalk between hypoxic and normoxic cells. To address this, we cultured
primary human adult cardiac fibroblasts in our three-compartment device, serum-arrested them
for 48h, and flowed 0% and 10% oxygen through the channels such that cells were cultured in
uniform normoxia, uniform hypoxia, or normoxia-hypoxia co-culture for 4h or 24h. First, we
immunostained tissues for hypoxia inducible factor 1α (HIF-1α), the most common member of the
HIF family of transcription factors238. As shown in Figure 3-3A-B, HIF-1α accumulated in the nuclei
of hypoxic fibroblasts by 4h, regardless of co-culture with normoxic cells. At 24h, HIF nuclear

64
localization mostly subsided (Figure 3-4A-B). We then assessed proliferation by staining cells for
Ki-67. In all conditions, cells expressed similar proliferation rates of around 20% after 4h of oxygen
modulation (Figure 3-3C-D). The percentage of cells expressing Ki-67 then decreased to a similar
extent in all conditions after 24h of oxygen modulation (Figure 3-4C-D). Finally, we assessed
myofibroblast differentiation by immunostaining for α-smooth muscle actin and actin. After 4h and
24h of oxygen modulation, we found no differences in the coverage of α-smooth muscle actin
relative to actin between conditions (Figure 3-3E-F and 3-4E-F). Thus, myofibroblast
differentiation at the protein level seems unaffected by oxygen modulation at 4h and 24h
timepoints.
Figure 3-3. Figure 2. Co-Culture Between Hypoxic and Normoxic Human Cardiac Fibroblasts for
4h. Immunofluorescence of (A-B) HIF-1α (red), DAPI (blue), (C-D) Ki-67 (green), (E-F) α-SMA
(red), actin (green), and DAPI (blue). Error bars depict mean with standard deviation from n=3
independent trials. *p<0.05, according to ordinary one-way ANOVA with Tukey’s multiple
comparisons test. Brightness and contrast are adjusted on all images. Scale bars = 100 µm.

65
Figure 3-4. Co-Culture Between Hypoxic and Normoxic Cardiac Fibroblasts for 24h.
Immunofluorescence of (A-B) HIF-1α (red), DAPI (blue), (C-D) Ki-67 (green), (E-F) α-SMA (red),
actin (green), and DAPI (blue). Error bars depict mean with standard deviation from n=3
independent trials. *p<0.05 **p<0.01 ****p<0.0001, according to ordinary one-way ANOVA with
Tukey’s multiple comparisons test or the Kruskall-Wallis test. Brightness and contrast are
adjusted on all images. Scale bars = 100 µm. 
Hypoxia Activates HIF-Signaling and Reparative Phenotypes in Cardiac Fibroblasts
To more broadly identify phenotypic differences, we next lysed cardiac fibroblasts after 24
hours of normoxia, hypoxia, or normoxia-hypoxia co-culture and performed bulk RNA sequencing.
Principal Component Analysis revealed that approximately 32% of the variance across all
samples was attributed to the first principal component, which clusters all hypoxic samples apart
from all normoxic samples, regardless of co-culture (Figure 3-5A). We detected 1949 differentially
expressed genes (DEGs) between hypoxia and normoxia and 2038 DEGs between hypoxic coculture and normoxia. Of these genes, 1486 were shared by both comparisons (Figure 3-5B),

66
indicating a relatively high amount of overlap. Thus, hypoxia was the most dominant regulator of
gene expression. Volcano plots for differentially expressed genes are shown in Figure 3-5C-D.
Pathway analysis on genes that were differentially expressed between uniform hypoxia
and uniform normoxia and/or between hypoxic co-culture and uniform normoxia indicated likely
upregulation of Glycolysis I, HIF-1α Signaling, signaling by VEGF, Gluconeogenesis I, and
Glycogen Metabolism (Figure 3-5E), which is expected in response to hypoxia. Significantly
upregulated genes represented in these pathways include the glycolytic enzymes ENO1 and
LDHA (Figure 3-5F). Several pathways related to fibrosis were also upregulated in both hypoxianormoxia comparisons, such as the Tumor Microenvironment Pathway, Wound Healing
Signaling, Collagen Biosynthesis and Modifying Enzymes, and Extracellular Matrix Organization
(Figure 3-5E), suggesting that fibrotic gene expression programs are activated by hypoxia.
Related to these pathways, hypoxia increased the expression of various mediators of wound
healing, including TGFB1 and VEGFA (Figure 3-5F). Hypoxic fibroblasts in both uniform and coculture conditions also upregulated anti-inflammatory pathways, such as signaling by Interleukin4 and -13 (Figure 3-5E), which are generally associated with pro-fibrotic phenotypes during wound
healing. Therefore, hypoxia shifted the transcriptome towards glycolysis and a pro-fibrotic
phenotype consistent with the anti-inflammatory phase of wound healing.
Some pathways that could have an impact on cardiac myocytes were also enriched in our
hypoxia versus normoxia comparisons. For example, we observed upregulation of Cardiac
Hypertrophy Signaling and β-Adrenergic Signaling in both comparisons (Figure 3-5E), which may
impact cardiac myocyte force generation. Beta blockers, which inhibit β-adrenergic receptors, are
commonly prescribed following infarction and are associated with reduced mortality
10
.
Interestingly, activation of β-adrenergic receptors in cardiac fibroblasts has been shown to drive
cell proliferation239
.

67
Figure 3-5. Oxygen-Dependent Changes in Gene Expression in Human Cardiac Fibroblasts. (A)
Principal Component Analysis of RNA-sequencing data. (B) Venn diagram of differentially
expressed genes in all conditions versus normoxia. (C-D) Volcano plots and (E) pathway analysis
dot plots for shared pathways between hypoxia versus normoxia and hypoxic co-culture versus
normoxia. (F) Normalized expression of genes related glycolysis and wound healing. Error bars
reflect mean with standard deviation of three pooled samples across n=6 independent trials.
***p<0.0005 and ****p<0.0001, according to ordinary one-way ANOVA with Tukey’s multiple
comparison’s test.

68
3.3.3. Co-Culture with Normoxic Fibroblasts Activates Pro-Inflammatory Phenotypes in Hypoxic
Fibroblasts
To identify transcriptional differences unique to cells in hypoxic-normoxic co-culture, we
compared cells in the normoxic co-culture region to cells in normoxia (Figure 3-6A) and cells in
the hypoxic co-culture region to cells in hypoxia (Fig 3-6B). Because these comparisons had fewer
differentially expressed genes, we removed the false discovery rate cutoff but applied a
strengthened p-value cutoff of 0.01. When comparing normoxic co-culture to uniform normoxia,
we detected 91 DEGs. Pathway analysis showed that these DEGs result in few coordinated
changes in pathways that are enriched but not associated with z scores, including Gap Junction
Signaling and Endothelin-1 Signaling. Of note, we observed antisense genes associated with
glycolysis (ENOX1-AS1) and myofibroblast activation (ACTA2-AS1) were significantly different in
normoxic co-culture versus control conditions (Fig 3-6C). Expression of these molecules changed
to levels comparable to hypoxic cells.
When comparing hypoxic co-culture to uniform hypoxia, we detected 122 DEGs. Pathway
analysis revealed that these are associated with the activation of several immune cell signaling
pathways, including Macrophage Classical Activation and Interferon Gamma Signaling (Fig 3-
6B). These pathways are associated with activation of the pro-inflammatory macrophage
phenotype that occurs in early wound healing240. We also observed upregulation of the effector
molecule PTX3, which was originally identified as a tumor-necrosis factor inducible gene in
fibroblasts241 and is now known to be induced by pro-inflammatory cytokines in a variety of cell
types242
. PTX3 serum levels are also elevated post-myocardial infarction243 and in heart failure244
.
We also see upregulation of the pro-inflammatory chemokine CCL5, which is involved in recruiting
immune cells to sites of inflammation245 (Fig 3-6C). Thus, compared to cells in uniform hypoxia,
hypoxic cells that are co-cultured with normoxic cells up-regulate several pathways and genes
related to pro-inflammatory signaling.

69
Next, to assess signaling molecules that may be acting up- or down-stream of proinflammatory gene expression, we measured an array of cytokines in cell media harvested from
all three compartments after 24h of conditioning on the device. Interestingly, tumor necrosis
factor-alpha (TNF-α) was uniquely upregulated in the co-culture media relative to both hypoxic
and normoxic media (Figure 3-6D). TNF-α is also elevated in serum following myocardial
infarction246, 247 and serum levels are correlated with heart failure247 and recurrent infarction246
. All
other cytokines measured did not show differences between conditions (Figure 3-7)

70
Figure 3-6. Crosstalk-Dependent Changes in Gene Expression in Human Cardiac Fibroblasts.
Volcano plots and pathway analysis of differentially expressed genes in (A) normoxic co-culture
versus normoxia or (B) hypoxic co-culture versus hypoxia. (C) Normalized counts of genes
significant to co-culture. (D) Pro-inflammatory concentrations of IFN-γ and TNF-α in cell media
isolated from hypoxia, hypoxia-normoxia co-culture, or normoxia regions. Error bars depict mean
with standard deviation from three pooled samples across n=6 independent trials. *p<0.05
**p<0.01, according to ordinary one-way ANOVA with Tukey’s multiple comparisons test.

71
Figure 3-7. Cytokine Levels in Media Isolated from Hypoxic, Normoxic, or Co-Culture Fibroblasts.
(A) Pro-inflammatory and (B) anti-inflammatory cytokine levels in media isolated from uniform
hypoxia, uniform normoxia, or hypoxia-normoxia co-culture regions of the device after 24h using
the MSD Human V-Plex Pro-Inflammatory 10-Plex Assay Kit. Statistical testing was performed
according to ordinary one-way ANOVA with Tukey’s multiple comparisons test or the KruskalWallis test.
3.4. Discussion
Myocardial infarctions cause injury by locally depriving tissue of oxygenated blood, which
initiates a complex repair process that is a dynamic interplay between many different cell types
within healthy and injured tissue. For example, immune cells, fibroblasts, and even border zone
cardiac myocytes have been shown to secrete cytokines and chemokines that may orchestrate
post-infarct wound healing, but the relative contribution of different cell types is not known29 and
very challenging to tease apart in vivo, limiting our ability to prevent pathological inflammation or

72
fibrosis post-MI. To gain a deeper understanding of the post-MI wound healing process, we
developed microfluidic devices to spatiotemporally control oxygen levels in vitro, similar to
previous studies70, 233, 235. We fabricated our devices from 3-D printed templates which enables
cost-effective, high-throughput, and rapid prototyping of microfluidic networks.
After validating that our device applied oxygen as expected, we then implemented it to coculture human cardiac fibroblasts experiencing 10% oxygen with those experiencing 0% oxygen,
mimicking healthy and infarcted myocardium, respectively. Immunostaining revealed that hypoxic
fibroblasts express HIF within 4h and that both hypoxia and physiological normoxia inhibit
proliferation after 24h to the same extent. Compared to ambient oxygen, hypoxia has been shown
to decrease proliferation of most cell types187 and HIF-1 specifically has been shown to suppress
proliferation of cardiac fibroblasts in a mouse model of infarction181. Although immunostaining
data did not show increased expression of α-SMA by 24h, we did observe increased expression
of fibrotic genes and pathways through RNA sequencing. This may indicate that changes on the
protein level require longer than 24h to be detectable. One study with human cardiac fibroblasts
saw increased α-SMA after 4-8 days of exposure to 1% O2
51 and in vivo, resident fibroblasts do
not express α-SMA until the proliferative phase of wound healing, which begins day 3 post-MI248
.
Similar to the immunostaining data, RNA sequencing revealed that hypoxia has a
profound effect on cell phenotype, with hypoxic cells shifting towards glycolysis and a more fibrotic
phenotype by 24h. We saw increased expression of TGFB1, a fibrotic growth factor which has
pleiotropic effects in the infarct border zone, including suppression of immune cells, activation of
fibroblasts into myofibroblasts, and hypertrophy of cardiac myocytes249
. We also saw increased
VEGFA, which is up-regulated after an infarction to drive angiogenesis250 and LOX, involved in
the cross-linking of collagen, which contributes to cardiac dysfunction post-infarction251
. We also
saw increased anti-inflammatory signaling pathways, such as IL-4 and -13, in hypoxic cells, which
are commonly associated with tissue repair and fibrosis. Hypoxic fibroblasts also upregulated
pathways that may impact cardiomyocyte function, including cardiac hypertrophy signaling and

73
β-Adrenergic signaling, which may promote compensatory mechanisms in neighboring cardiac
myocytes after injury. Dysregulation of these processes by hypoxia could be further investigated
and possibly exploited as targets for anti-fibrotic therapies.
Importantly, RNA-Seq of tissues and analysis of cell media demonstrated that, when
cultured with normoxic fibroblasts, the hypoxic co-culture region upregulates pro-inflammatory
signaling relative to hypoxic controls, such as macrophage classical activation signaling,
expression of the chemokine CCL5, and secretion of TNF-α. Previous studies have also shown
that TNF-α secretion by human cardiac fibroblasts is not induced by uniform hypoxia170, 252
,
consistent with our finding that this phenotype was unique to the hypoxia-normoxia crosstalk.
Because TNF-α is a potent inflammatory signal upregulated post-MI246, 247, its exacerbation
through hypoxia-normoxia crosstalk should be explored further. In contrast, normoxic cells in coculture with hypoxic cells alter antisense genes related to glycolysis and fibrosis, to levels
comparable to hypoxic cells. Antisense RNA molecules can regulate gene expression, generally
inhibiting translation to protein. Therefore, dysregulation of these antisense genes in normoxic
co-culture may be potentially explored as anti-fibrotic targets.
To improve upon conventional in vivo animal models, which display the complex oxygen
landscapes of post-infarct myocardium but lack human-relevance, we chose to incorporate
human adult cardiac fibroblasts in our devices. It is important to note that this is a primary cell
source that comes from a single female patient, and therefore limits its relevance to other patient
responses. However, the use of a patient-specific cells demonstrates the ability for this platform
to be implemented to study personalized responses to disease or drug treatments, especially with
the advent of human induced pluripotent stem-cell derived cardiac cells. Together, this work
identifies both oxygen-dependent and crosstalk-dependent changes in cardiac fibroblast
phenotype that are critical for understanding and mitigating pathological fibrosis following
infarction.

74
Chapter 4 A Post-Infarct Human Myocardium on a Chip Reveals Hypoxic Cardiac
Fibroblasts Drive Cytokine Signaling in Normoxic Cardiac Myocytes
This Chapter is an adaptation of: Khalil Natalie N., Rexius-Hall Megan L., Gupta Divya, McCarthy
Liam, Verma Riya, Kellogg Austin C., Takamoto Kaelyn, Xu Maryann, Nejatpoor Tiana, Parker
Sarah J., and McCain Megan L. Crosstalk with Hypoxic Cardiac Fibroblasts Activates ProInflammatory Signaling in a Post-Infarct Human Myocardium on a Chip. Advanced Healthcare. In
Review.
In the wound healing process that follows a myocardial infarction, hypoxic and normoxic
cardiac cells cooperate to replace injured tissue with a scar. Cardiac fibroblasts play a critical role
in this process by activating into a myofibroblast phenotype, which synthesizes and secretes
matrix proteins. Cardiac fibroblasts are also a prolific source of cytokines that may impact
neighboring cardiac cell types. In this chapter, we co-culture hypoxic fibroblasts with normoxic
human induced pluripotent stem cell (hiPSC)-derived cardiac myocytes to identify crosstalkdependent cardiac cell phenotypes. We robustly characterize phenotype through
immunofluorescence, RNA-sequencing, and analysis of cytokine secretion. We find that coculture increased many pathways associated with cytokine signaling, including upregulation of
the IL-6 family signaling pathway and increased expression of the IL-6 receptor in normoxic
myocytes. Consistent with this, hypoxic fibroblasts secreted higher levels of an array of several
cytokines, including IL-6, than myocytes. Together, these findings suggest that hypoxia-normoxia
crosstalk between cardiac cell types upregulates inflammatory signaling pathways resembling the
initial inflammatory phase of wound healing.
4.1. Introduction
As described, myocardial infarctions deprive downstream tissue of oxygenated blood. The
resulting hypoxic injury is followed by a repair process that necessitates cooperation between
hypoxic tissue of the infarct zone and neighboring healthy tissue. For example, immune cells are
first recruited into the infarct zone to clear dead cells and matrix debris, which is facilitated by pro-

75
inflammatory cytokines and chemokines, such as TNF-α and IL-6
28
. Timely repression of this
inflammation is critical to mitigate the extent of injury29
. As immune cells ultimately transition
towards resolving phenotypes, they secrete anti-inflammatory mediators, such as TGF-β1, that
drive cardiac fibroblasts to proliferate and activate into a myofibroblast phenotype. The
myofibroblast is responsible for depositing matrix proteins that ultimately form a scar during
wound healing30. While some level of scarring may be necessary to maintain tissue integrity253
,
fibrosis is ultimately a risk factor for heart failure254
. The mechanisms that drive unresolved fibrosis
or inflammation are incompletely understood and thus difficult to prevent.
Because healthy and injured myocardium are in constant communication during wound
healing, we hypothesized that crosstalk between hypoxic fibroblasts and neighboring, normoxic
cardiac cell types may regulate mechanisms of inflammation or fibrosis, This has not fully been
explored due to limitations of conventional research tools. Conventional research has typically
relied on the use of a hypoxia chamber, which is limited to studying a single, global oxygen level.
To then explore crosstalk, studies have relied on the transfer of medium conditioned by one cell
type after a fixed duration of hypoxia. These studies have demonstrated that hypoxic fibroblasts
reduce normoxic cardiac myocyte viability61, 64. However, these studies have relied on non-human
cell sources and signaling factors in spent media. They also typically culture cells in ambient, 21%
oxygen, as the reference for normoxia, which is much higher than physiological 10% oxygen that
has been measured in remote zones of post-infarct myocardium34, 35
. Thus, they do not reflect
real-time cell-to-cell signaling, which includes the spatiotemporal dynamics of signaling
molecules, such as feedback loops between cells. To address this, microfluidic devices for oxygen
control have emerged which enable the real-time co-culture of hypoxic and normoxic cardiac
cells70, 233
. In Chapter 3, we developed microfluidic devices from 3D-printed templates to examine
oxygen-dependent crosstalk between hypoxic and normoxic human adult cardiac fibroblasts. We
found that inflammatory signaling was augmented in cardiac fibroblasts after hypoxia-normoxia
crosstalk. Here, we re-engineered similar devices for the co-culture of hypoxic human cardiac

76
fibroblasts with normoxic human-induced pluripotent stem cell (hiPSC)-derived cardiac myocytes
to recapitulate multicellular crosstalk in post-infarct myocardium. We find that co-culture with
hypoxic fibroblasts results in reduced cardiac myocytes cell counts and altered cell morphology,
suggestive of impaired viability. RNA-Seq analysis of co-culture tissues demonstrates increases
in pathways related to cytokine signaling, including IL-6-type family singlaing, which was
accompanied by an increase in gene expression of the IL-6 receptor and STAT3 in cardiac
myocytes. Consistent with this, analysis of cell media demonstrates that hypoxic cardiac
fibroblasts are a more prolific source of several cytokines relative to hiPSC-derived cardiac
myocytes, which may be facilitating the increased cytokine signaling observed in co-culture.
4.2. Materials and Methods
4.2.1. Microfluidic Device Fabrication
Microfluidic devices with channels for gas flow were designed and validated as described
in Chapter 2. Briefly, channel networks were made in Autodesk Inventor (Autodesk, Inc., San
Francisco, CA, USA), 3D-printed using the μMicrofluidics Edition M-50 3D Printer with Master
Mold Photopolymer Resin (CADworks3D, Concord, ON, Canada), and replica molded with
polydimethylsiloxane (PDMS). PDMS was prepared by mixing base and curing agent of a Sylgard
184 Silicone Elastomer kit (Dow, Midland, MI, USA) in a 10:1 weight ratio in a planetary centrifugal
mixer (AR-100, Thinky, Japan). Replicas were then biopsy punched and channels were sealed
by bonding to 100-µm PDMS membranes, fabricated by spin-coating thin layers of PDMS onto
blank silicon wafers using predefined spin settings235 on a G3P-8 spin-coater (Specialty Coating
Systems, Indianapolis, IN, USA). Finally, sealed microfluidic devices were bonded to Cell Culture
Chambers to enclose cells and media.

77
4.2.2. Cell Culture Chamber and Co-Seeding Chamber Fabrication
Cell Culture Chambers to multiplex the device into three cell culture regions, as well as
Co-Seeding Chambers to enable dual-seeding in the central cell culture region, were designed
using Autodesk Inventor (Autodesk, Inc., San Francisco, CA, USA). CAD designs were 3D-printed
in acetonitrile butadiene (ABS) or polylactic acid (PLA) on a FlashForge Creator Max 2
Independent Dual Extruder 3D Printer (FlashForge USA, Industry, CA, USA) and then replicamolded with PDMS. PDMS was cured for 4h at 65°C, after which molds were soaked in acetone
to remove the replica. Similar to work in Chapter 2, Cell Culture Chamber replicas were then
enclosed with a lid by painting with uncured PDMS and placing in contact with a 5-mm thick PDMS
slab that was cut to size with a razor blade.
4.2.3. Dual-Seeding
PDMS lids were peeled off of devices for UVO-treating, cell seeding, and media changes,
then re-attached again with uncured PDMS the day of experiments. Devices were UVO-treated
for 8 minutes and then uniformly coated with 150 µL of human fibronectin (50 µg/mL, Corning,
Corning, NY, USA) for 1 hour. Residual fibronectin was removed with compressed air and devices
were left to dry in the biosafety cabinet for at least 1 hour to ensure a dry surface for optimal
water-tight contact with the Co-Seeding Chamber. Co-Seeding Chambers were UVO-treated for
8 minutes and then placed in contact with dried, fibronectin-coated PDMS devices to form a watertight seal. Cells were added to each region as described in cell culture methods.
4.2.4. Human Cardiac Fibroblast Cell Culture
Primary human adult cardiac fibroblasts (Lot 3131, Cell Applications Inc., San Diego, CA,
USA) were thawed into a 75 cm2 cell culture flask and cultured in low glucose DMEM (1g/L
glucose) supplemented with 10% v/v fetal bovine serum (Gibco, Waltham, MA, USA) and 1% v/v

78
penicillin-streptomycin. Cells were passaged in a 1:10 ratio at 70-80% confluence into new 75
cm2 cell culture flasks, by incubating cells with trypsin-EDTA for 5 minutes at 37°C and
neutralizing with 10% serum. For experiments, cells at passages four through eight were seeded
onto PDMS devices with fibronectin at a density of 22,500 cells/cm2
.
4.2.5. Human iPSC-derived Cardiac Myocyte Differentiation and Cell Culture
Control human iPSCs were gifted from the laboratory of Dr. Justin Ichida and were
reprogrammed from lymphocytes acquired from the NINDS Biorepository (ND05280, Coriell
Institute for Medical Research, Camden, NJ, USA), as described previously255, 256. iPSCs were
maintained on 6-well plates coated with Matrigel (1:100; Corning, Glendale, AZ, USA) and
mTESR plus media (STEMCELL Technologies, Vancouver, BC, Canada). At 80% confluence,
iPSCs were plated on Matrigel-coated plates using Accutase (STEMCELL Technologies,
Vancouver, BC, Canada) at a seeding density of 1.5 x 106 cells per well of a 6-well plate. Cells
were in mTESR plus media supplemented with 10 µM ROCK inhibitor (Y-27632; Selleck
Chemicals, Houston, TX, USA) for the first 24h. To recover iPSC morphology, cells were
passaged for minimum of three times prior to differentiation using 0.5 mM EDTA. IPSCs were
then differentiated according to the STEMdiff Ventricular Cardiomyocyte Differentiation Kit
(STEMCELL Technologies, Vancouver, BC, Canada). At day 10, cells were placed in purification
media for two days composed of RPMI 1640 Medium, no glucose (Thermo Fisher Scientific,
Waltham, MA, USA) and B-27 Supplement, minus vitamin A (Thermo Fisher Scientific, Waltham,
MA, USA). At day 12, the cells were placed in fatty acid media composed of RPMI 1640 Medium,
no glucose (Thermo Fisher Scientific, Waltham, MA, USA) supplemented with L-carnitine, and
BSA conjugated with Linoleic Acid, Oleic Acid, and Palmitic Acid according to a published
protocol257. Cells were then seeded onto devices to achieve 250,000 cells per compartment.

79
4.2.6. Oxygen Modulation
The day after seeding, the co-culture chamber was removed, a media change with 0.1%
FBS low glucose DMEM was provided to all regions of the device, and devices were bonded to
lids using uncured PDMS. Microfluidic devices were connected to tubing from 0% O2 (5% CO2,
bal N2) and 10% O2 (10% O2, 5% CO2, bal N2) compressed gas cylinders, as described
previously70. Pressure in the two microchannels was kept equal at 4 psi similar to previous work
in Chapter 3.
4.2.7. Immunostaining and Image Analysis
Cells were fixed with 4% paraformaldehyde for 10 minutes then permeabilized with 0.5%
Triton X-100 in PBS for 10 minutes. Cells were then incubated with a primary staining solution at
room temperature for 1 hour containing monoclonal rabbit vimentin (1:200, Novus Biologicals,
Littleton, CO, USA) and monoclonal mouse alpha actinin (1:200, Thermo Fisher Scientific,
Waltham, MA, USA) in a 10% Normal Goat Serum (Thermo Fisher Scientific, Waltham, MA, USA).
Cells were rinsed with PBS and then incubated at room temperature for 1.5h with Alexa Fluor 547
goat anti-mouse secondary antibody (1:200, Manufacturer), Alexa Fluor 547 goat anti-rabbit 488
(1:200, manufacturer), DAPI (1:200, Life Technologies, Waltham, MA, USA), or Alexa Fluor 488
Phalloidin (1:200, Life Technologies). Cells were mounted onto glass slides using Prolong Gold
Anti-Fade (Life Technologies).
4.2.8. RNA and Protein Isolation
Coverslips were incubated in TRIzol immediately after device experiments to lyse and
homogenize cells. TRIzol-chloroform phase separation was used for the extraction of mRNA from
the aqueous phase. mRNA transcripts were isolated using a Qiagen miRNA Mini Kit (Bio-Rad,
Hercules, CA, USA).

80
4.2.9. RNA-Sequencing and Data Analysis
Each sample used for RNA-Sequencing was pooled from two replicates across four
independent trials. RNA was sequenced as described in Chapter 3. Briefly, RNA integrity was
assessed using 2100 Bioanalyzer using the Agilent RNA 6000 Nano Kit (Agilent Technologies,
Santa Clara, CA) and total RNA was quantified using a Qubit RNA HS Assay Kit (ThermoFisher
Scientific, Waltham, MA). Samples were purified for mRNA using the NEBNext® Poly(A) mRNA
Magnetic Isolation Module (New England Biolab Inc, Ipswich, MA). Libraries were constructed
using the xGen Broad-Range RNA Library Prep Kit (Integrated DNA Technologies, Coralville, IA).
RNA-sequencing was performed using the NovaSeq 6000 platform (Illumina, San Diego, CA) with
75bp single-end sequencing.
As described in Chapter 3, sequences were analyzed using Partek Flow Genomic
Analysis Software (Partek Inc., St. Louis, MO, USA). Sequences were trimmed for using a Phred
score of 20 and min read length 25, aligned to the hg38 genome using STAR (2.7.8a), and
quantified using encode 38 (V2). Counts underwent median ratio normalization using DESeq2
method. Differentially expressed genes that met fold change and false discovery rate cutoffs of
1.5, and 0.05, respectively, were used in Ingenuity Pathway Analysis (QIAGEN, Hilden,
Germany). Data have been made publicly available at Accession No. GSE254962 on Gene
Expression Omnibus.
4.2.10. MSD Assay
Cytokine secretion was analyzed in cell media isolates using the MSD Human V-Plex ProInflammatory 10-Plex Assay Kit (Meso Scale Diagnostics, LLC., Rockville, MD, USA) and MSD
QuickPlex SQ 120MM Imager (Meso Scale Diagnostics, LLC., Rockville, MD, USA), according to
the manufacturer’s protocol. Each sample was a pooling from two replicates across n=4
independent trials.

81
4.3. Results
4.3.1. A 3D-Printed and Replica-Molded Chamber Enables Dual Seeding of Human
Cardiac Fibroblasts and Human iPSC-Derived Cardiac Myocytes
We previously developed microfluidic devices that enable the real-time co-culture of
hypoxic and normoxic cardiac cells235. Briefly, cells are cultured on top of a thin, gas-permeable
membrane located above two microchannels for gas flow. Gases are perfused through the
channel network and rapidly diffuse upwards across the membrane to reach cells. We perfused
10% and 0% oxygen (O2) gas through the two microchannels, which correspond to oxygen levels
in healthy and infarcted myocardium35, respectively. Microfluidic devices are bonded to a Cell
Culture Chamber with three compartments that generates internal normoxia and hypoxia control
regions (Fig 4-1A). Microfluidic devices are fabricated through DLP 3D-printing and replica
molding in polydimethylsiloxane, a transparent and gas-permeable elastomer, as described in
Chapter 2.
Here, we re-engineered these devices to enable the co-culture of two distinct cell types.
To do this, we 3D-printed and replica molded a PDMS Co-Seeding Chamber that is placed on the
device surface prior to cell seeding, and then removed once cell types are adherent (Figure 4-
1B). Using this method, we were able to co-culture primary human adult cardiac fibroblasts
adjacent to human induced pluripotent stem cell (iPSC)-derived cardiac myocytes in the central
co-culture region of the device, as demonstrated with brightfield images (Fig 4-2A).

82
Figure 4-1. Microfluidic Devices Adapted to Co-Culture Hypoxic Cardiac Fibroblasts with
Normoxic Cardiac Myocytes. (A) Schematic of microfluidic devices, which contain two channels
for hypoxic and normoxic gas flow. Gases diffuse across the cell culture membrane to reach cells
cultured on the surface. Devices are bonded to a cell culture chamber which multiplexes the
device into a hypoxia, normoxia, and co-culture region. (B) Flow chart depicting use of a CoSeeding Chamber to enable seeding of two distinct cell types on microfluidic devices.
4.3.2. Hypoxic Cardiac Fibroblasts Impair Neighboring Normoxic Cardiac Myocyte Cell
Counts and Morphology
After removal of the dual-seeding chamber, we immediately perfused 0% oxygen beneath
fibroblasts and 10% oxygen beneath myocytes for 24h. We stained devices for DAPI as well as
α-actinin and vimentin to detect cardiac myocytes and fibroblasts, respectively. As shown in
Figure 5B, α-actinin signal was restricted to the cardiac myocyte half of the co-culture region,
demonstrating successful separation of cell types. However, cardiac myocytes appear to express
some level of the non-myocyte marker vimentin, perhaps due to their inherently immature cell

83
phenotype. Vimentin is expressed by fetal cardiac myocytes258 and hiPSC-derived cardiac
myocytes259 and diminishes with increasing fetal age258
.
Qualitatively, cardiac myocytes appeared smaller and more aggregated than cardiac
fibroblasts. We also observed fewer cardiac myocytes in co-culture with hypoxic fibroblasts
relative to control cardiac myocytes, which was confirmed by quantifying nuclei size and number
(Figures 5C-D). Reduced cell counts may be suggestive of decreased cell viability after co-culture
with hypoxic fibroblasts, consistent with previous work that demonstrated that medium from
hypoxic fibroblasts impairs normoxic cardiac myocyte viability61, 64
.

84
Figure 4-2. Co-Culture of Normoxic hiPSC-derived Cardiac Myocytes and Hypoxic Cardiac
Fibroblasts for 24h. (A) Brightfield of cardiac fibroblasts (CFs) and myocytes (CMs) 12h after cell
seeding. Microfluidic channels can be visualized beneath the cell culture membrane. (B)
Fluorescence microscopy of DAPI (blue), vimentin (green), and α-actinin (red) in hypoxic CFs and
normoxic CMs in control or co-culture regions for 24h. Brightness and contrast are adjusted on
all images. Scale bars = 500 µm. Quantification of nuclei (C) size and (D) number using the
ImageJ analyze particles function. Error bars depict mean with standard deviation from n=3
independent trials. (E) Total RNA concentration from samples used for RNA-sequencing. Error

85
bars depict mean with standard deviation from two pooled replicates across n=4 independent
trials. * p < 0.05, *** p < 0.0005, and **** p < 0.0001 according to ordinary one-way ANOVA with
Tukey’s multiple comparisons test.
4.3.3. RNA-Sequencing Demonstrates Primary Clustering by Cell Type
We next isolated and sequenced RNA from the co-culture and control regions after 24h
on the device. Total RNA extracted from our samples mirrored the cell count data (Figure 4-2E),
suggesting fewer cardiac myocytes in co-culture with hypoxic fibroblasts. To first visualize global
differences in the transcriptome, we performed principal component analysis and hierarchical
clustering. Both analyses demonstrate primary clustering of hypoxic fibroblasts away from
normoxic cardiac myocytes, regardless of co-culture (Fig 4-3A-B). This is expected because of
the stark differences in gene expression between different cell types.
Figure 4-3. The Global Transcriptome of Cardiac Fibroblasts (CFs) and Cardiac Myocytes (CMs)
in Hypoxia, Normoxia, or Co-Culture for 24h. (A) Principal component analysis and (B)
hierarchical clustering on RNA-sequencing data demonstrates primary clustering of samples
based on oxygen-level and cell type, regardless of co-culture. Each sample reflects a pooling of
two samples across four independent trials.

86
4.3.4. Normoxic Cardiac Myocytes Increase Cytokine Signaling After Co-Culture with
Hypoxic Cardiac Fibroblasts
We next isolated and sequenced RNA from all regions of the device after 24h. Total RNA
extracted from our samples mirrored the cell count data (Figure 4-3E), suggesting fewer cardiac
myocytes in co-culture with hypoxic fibroblasts. To visualize global differences in the
transcriptome, we again performed Principal Component Analysis. PCA plots demonstrate nearly
87% of the variance is described by PC1, which clusters hypoxic fibroblasts away from normoxic
cardiac myocytes, regardless of co-culture (Figure 4-4A). This is expected because of the stark
differences in gene expression between different cell types.
Differential analysis identified 573 DEGs between hypoxic fibroblasts and hypoxic
fibroblasts in co-culture, and 260 DEGs between normoxic cardiac myocytes and normoxic
myocytes in co-culture. Pathway analysis of DEGs in hypoxic fibroblasts co-cultured with
normoxic myocytes identified some genes and pathways normally associated with cardiac
myocytes, such as an increase in the striated muscle contraction pathway (Figure 4-4B). We think
these DEGs are an artifact caused by minor mixing of cell types in the co-culture chamber, which
may register as drastic changes in gene expression since fibroblasts in monoculture likely have
negligible expression of these genes. Other pathways observed in hypoxic fibroblasts in coculture include decreased RHOGDI signaling. As RHOGDI is an inhibitor of Rho GTPases, this
may suggest increased Rho activity, which are critical regulators of cell morphology and
migration260. Consistent with this, previous studies have shown paracrine factors from hypoxic
myocytes drive fibroblast migration59, 60. We also observed increased neurovascular coupling
signaling, which may suggest increased cell coupling in response to paracrine factors from
neighboring myocytes. Coupling between cardiac myocytes and fibroblasts has been observed in
other species215-217 and may contribute to the development of arrythmias.

87
In normoxic myocytes co-cultured with hypoxic fibroblasts, we observed upregulation of
various pathways involved in cytokine signaling, including Pathogen Induced Cytokine Storm
Signaling, IL-6 Family Signaling, Interferon Alpha/Beta Signaling, and Interferon Gamma
Signaling (Figure 4-4B). Thus, we again observed increased pro-inflammatory signaling pathways
after hypoxia-normoxia co-culture. Normoxic cardiac myocytes in co-culture with hypoxic
fibroblasts increased expression of genes involved in cytokine signaling, including IL-6, the IL-6
receptor (IL6R), the suppressor of cytokine signaling 3 (SOCS3), and STAT3 (Figure 4-4C).
Plasma levels of IL-6 in the acute phase post-MI are correlated with left ventricular remodeling261
and, consistent with this, inhibition of the IL-6 receptor has improved contractile function and
remodeling post-MI262
. Increased IL-6/STAT3 signaling has also been associated with cardiac
rupture263. Therefore, the augmented IL-6/STAT3 signaling we observed in hypoxia-normoxia coculture may be maladaptive and can be explored further as a therapeutic target.

88
Figure 4-4. Crosstalk-Dependent Changes in Gene Expression in Hypoxic Human Cardiac
Fibroblasts and Normoxic hiPSC-derived Cardiac Myocytes Co-Cultured for 24h. (A) Volcano
plots and (B) pathway analysis of differentially expressed genes between the central co-culture
and control regions. (C) Dot plots for co-culture specific genes and their expression in hypoxic
cardiac fibroblasts (CFs) and normoxic cardiac myocytes (CMs) in co-culture versus control
regions. Error bars reflect mean with standard deviation. Each sample is a pooling of two samples
across four independent trials. * p < 0.05, ** p < 0.005, *** p < 0.0005, and **** p < 0.0001
according to ordinary one-way ANOVA with Tukey’s multiple comparisons test or the KruskalWallis test.
4.3.5. Hypoxic cardiac fibroblasts are a more prolific source of cytokines than normoxic
cardiac myocytes
Finally, we measured an array of ten pro-inflammatory and anti-inflammatory cytokines in
cell media. Hypoxic fibroblasts secreted significantly higher levels of most cytokines, including IL6, IL-8, IL-1β, IL-12p70, IL-4 and IL-13, when compared to normoxic myocytes, which secreted
very low levels of all measured cytokines (Figure 4-5). All other cytokines (TNF-α, IL-10, IFN-γ)

89
approached significance (p < 0.15), with the exception of IL-2 (p > 0.999) (Figure 4-5). For the
most part, the co-culture compartment had levels of cytokines that were intermediate to hypoxic
fibroblasts and normoxic myocytes, suggesting a linear mixing of factors from both control
compartments. In the case of TNF-α, levels in hypoxia and co-culture regions were comparable
(Fig 4-5), suggesting it was uniquely upregulated in co-culture, similar to our observation with cocultured hypoxic-normoxic fibroblasts in Chapter 3. Together, these results echo previous work
that cardiac fibroblasts produce more cytokines than cardiac myocytes. While we did not see
differences in IL-6 secretion between hypoxic and normoxic fibroblasts, we did observe higher
secretion of IL-6 by fibroblasts than by cardiac myocytes, which has been demonstrated
previously264
. Fibroblasts have also been identified as the key source of IL-6 post-MI265 and are
also regarded as the main source of cytokines in inflammatory diseases, such as rheumatoid
arthritis266
.

90
Figure 4-5. Cytokine Secretion in Hypoxic Human Cardiac Fibroblasts and Normoxic hiPSCderived Cardiac Myocytes Co-Cultured for 24h. Mean concentrations of an array of ten (A) proand (B) anti-inflammatory cytokines in cell media. Error bars reflect mean with standard deviation
from two pooled samples across four independent trials. * p < 0.05, ** p < 0.005, *** p < 0.0005,
and **** p < 0.0001 according to ordinary one-way ANOVA with Tukey’s multiple comparisons
test or the Kruskal-Wallis test.

91
4.4. Conclusions
In this Chapter, we co-cultured hypoxic cardiac fibroblasts with normoxic hiPSC-derived
cardiac myocytes and found that co-culture with hypoxic fibroblasts reduced cardiac myocyte cell
count, suggestive of impaired cell viability. Hypoxia-normoxia crosstalk also again promoted proinflammatory signaling in normoxic cardiac myocytes, including increased gene expression of the
IL-6 receptor. Because hypoxic fibroblasts secreted higher levels of several pro- and antiinflammatory cytokines compared to normoxic myocytes, the changes in myocyte phenotype are
likely a direct result of secretions from hypoxic fibroblasts. Since we did not initially observe
differences between hypoxic and normoxic fibroblast cytokine production, this cytokine signaling
likely represents the constitutive levels of cytokines produced by fibroblasts. It is possible,
however, that there are other signaling molecules not measured in the pro-inflammatory panel
that may be induced by hypoxia in fibroblasts and may be impacting neighboring normoxic cells.
Together, these data suggest that hypoxia-normoxia crosstalk over acute timescales repeatedly
promoted pro-inflammatory pathways that resemble the initial inflammatory phase of wound
healing.
To preserve human relevance, we used human adult primary cardiac fibroblasts and
hiPSC-derived cardiac myocytes, which are both from a single (and different) donor. Human adult
primary cardiac fibroblasts are a relatively mature and physiologically-relevant cell source, but
may not translate to other patients’ responses and are difficult to source from multiple patients. In
contrast, hiPSC derivatives enable our type of platform to be implemented to identify patientspecific responses to disease or drug treatments in a less-invasive manner, with the tradeoff of
lower maturity. For example, hiPSC-derived cardiac myocytes in our study expressed the fetal
marker vimentin, similar to other studies259. Protocols to differentiate fibroblasts from hiPSCs have
also been developed, but these cells similarly resemble embryonic rather than adult

92
phenotypes267. Future work can involve maturing these cell types, for example through metabolic
maturation268-270 or electromechanical stimulation271, 272
.
Organ on a Chip models of myocardial infarction are increasingly reflecting the
spatiotemporal complexities of post-infarct myocardium232. These include models of hypoxia or
ischemia that harness gas-impermeable materials32, chemical hypoxia reagents65, inherent
diffusion limitations of spheroids84, and gas cylinders170, 235. Our approach to supply microfluidic
channels by cylinders of compressed gas enables precise control over oxygen at the microscale,
but has limitations in scalability because of the reliance on gas cylinder infrastructure. Nextgeneration Organ on a Chip models of myocardial infarction can focus on scalability and
incorporate more relevant tissue architectures, such as 3-dimensional cell-cell or cell-ECM
contacts150, 273. Furthermore, because our studies repeatedly demonstrated that fibroblasts play a
critical role in inflammatory cytokine signaling, future Organ on a Chip models can integrate
immune cells to explore fibroblast-myocyte-immune cell interactions in the context of hypoxia.
Combined with the advent of more hiPSC-derived cardiac cell types, scalable and physiologicallyrelevant Organ on Chip models of myocardial infarction will continue to evolve and serve as
advanced preclinical model systems for identifying mechanisms of pathophysiology and
screening patient-specific drug responses.

93
Chapter 5 Outlook
In this dissertation, we first discussed Organ on a Chip approaches to model the dynamic
cellular microenvironment of post-infarct myocardium. We next recapitulated the temporal
dynamics of acute hypoxic injury in vitro by characterizing cardiac fibroblast phenotypes after
distinct durations of hypoxia. We then developed a new approach to mimic the spatiotemporal
changes in oxygen that bridge healthy and injured tissue after a heart attack. We implemented
this to co-culture hypoxic (0% O2) cardiac fibroblasts with neighboring, normoxic (10% O2) cardiac
fibroblasts or hiPSC-derived cardiac myocytes. Together, these studies revealed new insights
into oxygen-dependent or crosstalk-dependent regulation of cardiac cell phenotypes that may
underlie mechanisms of the wound healing process post-infarction.
These chapters repeatedly demonstrated that hypoxic fibroblasts (0% O2) exhibited
nuclear translocation of HIF-1α, a key transcription factor that mediates the cellular response to
hypoxia, on acute timescales. Hypoxia also repeatedly inhibited the proliferation of human cardiac
fibroblasts, perhaps to conserve energy in the face of a limited metabolic resource. Previous
studies in mouse have demonstrated that proliferation of cardiac fibroblasts post-infarction is
suppressed by HIF-1α181
. Surprisingly, hypoxic fibroblasts globally repressed protein synthesis,
including via downregulation of major pathways impacting translation such as eukaryotic
translation initiation. We suspect that these phenotypes may reflect a reprogramming at the
protein level as cells are switching from proliferating to differentiating phenotypes, which has been
observed in other cell types, including hepatocytes274, adipose-derived stromal cells275, and CD8+
effector T cells212
. Consistent with this hypothesis, microfluidic oxygen control demonstrated that
the global transcriptome of hypoxic fibroblasts shift towards a more fibrotic phenotype, suggestive
of cell differentiation, which may perhaps be expressed at the protein level at later time points.
We next examined the impact of co-culture between hypoxic fibroblasts and other cardiac
cell types. When hypoxic fibroblasts were co-cultured with either normoxic fibroblasts or normoxic

94
hiPSC-derived cardiac myocytes, we repeatedly observed increased inflammatory signaling. In
hypoxic fibroblasts co-cultured with normoxic fibroblasts, we saw increased gene expression of
the chemokine CCL5, involved in recruiting and activating leukocytes, and an increase in the
macrophage classical activation pathway. Consistent with this, we saw increased secretion of the
cytokine TNF-α in co-culture. In normoxic cardiac myocytes co-cultured with hypoxic fibroblasts,
we observed upregulation of IL-6-family signaling, accompanied by increased gene expression of
the IL-6 receptor and STAT3. Hypoxic fibroblasts secreted significantly higher levels of nearly
every measured cytokine, including IL-6, than normoxic myocytes. Together, these results
implicate fibroblasts in early cytokine signaling under hypoxia-normoxia crosstalk that may impact
the initial inflammatory phase of wound healing. The mechanisms that might underlie this
increased inflammatory signaling after hypoxia-normoxia co-culture should be further explored.
In this outlook, we discuss both the biological and technological outlook for this work, including
how fibroblasts may be leveraged as therapeutic targets and future opportunities for post-infarct
myocardia on chips.
5.1. Fibroblasts as Therapeutic Targets
In Chapters 3 and 4, we found that hypoxic fibroblasts exacerbate inflammatory signaling
when cultured with normoxic fibroblasts or myocytes. Specifically, we saw increased TNF-α
secretion in co-culture with fibroblasts, and increased IL-6 type signaling in co-culture with
myocytes. Both TNF-α
246, 247 and IL-6
261, 263 are upregulated post-infarction and correlated with
disease severity, such as heart failure. As a result, both TNF-α and IL-6 inhibitors have been
pursued as pharmacological treatments post-infarction or in heart failure patients, but these have
mixed outcomes and potential for off-target effects. Briefly, although TNF-α is correlated with heart
failure, TNF inhibition does not improve, and may even adversely affect, patient outcome276, 277
.
One potential reason for this has been elucidated in a recent mouse model of myocarditis, where
TNF-α, despite its conventional pro-inflammatory role, also had anti-inflammatory effects through

95
inducing activation-induced cell death of heart-reactive T cells278
. Conversely, TNF-α inhibition
has reduced risk of infarction in patients with psoriasis279 or rheumatoid arhritis280, though these
patients have elevated risk at baseline due to systemic inflammation. IL-6 inhibitors, on the other
hand, have been effective in improving myocardial salvage and reducing neutrophil counts in
patients who have had myocardial infarction281, 282, though as with other anti-inflammatory
therapeutics, there is the potential for off-target effects, including increased risk of infection or
sepsis283, 284
.
To address this, more specific approaches can be pursued that target the maladaptive cell
type, such as the myofibroblast, rather than its ubiquitous signaling factor. Targeting fibroblasts,
which drive pathological remodeling, has been challenging because there is no single and unique
marker for them. This is partly due to their heterogeneity of cell origin – resident fibroblasts
primarily originate from the epicardium285 during development, although a fraction also arises from
endothelial286, 287 and Pax3-expressing cells287. In response to injury, endothelial-mesenchymal
transition has also been identified to contribute to the population of activated fibroblasts288
.
Despite this heterogeneity, one cell surface receptor that has been successful in targeting
myofibroblasts is fibroblast activation protein (FAP). FAP is expressed on the surface of activated
fibroblasts, including cardiac fibroblasts post-infarction289
, but also non-fibroblasts, such as
macrophages290 or epithelial cancers291, 292
. Despite this, CAR T cells have recently been
developed that can successfully target activated cardiac fibroblasts through expression of FAP.
FAP CAR T cells were shown to improve remodeling and cardiac function in a mouse model of
infarction293. However, CAR T cells can be associated with adverse events such as cytokine
release syndrome, a form of systemic inflammation294. As CAR T cell therapy advances, so will
its application in the context of cardiac fibrosis.
In the future, nanoparticle approaches can also be employed that leverage the same
receptor for local administration of therapeutics that target fibroblasts. These can deliver cargo
for gene editing in activated fibroblasts, for example to inhibit cytokine signaling, or can distribute

96
IL-6 and TNF-α antagonists locally in order to mitigate off-target effects. Nanoparticles for the
treatment of myocardial infarction295 have already been developed that improve cardiac function
in animal models, including through anti-inflammatory cargo296, 297
. More preventative approaches
can include targeting the atherosclerosis that leads to infarction298, for example by reducing the
size299 or increasing the stability300 of plaque buildup.
5.2. Towards a 4-D Human Myocardium on a Chip
In this dissertation, we developed a microfluidic device from 3D-printed templates for
spatiotemporal control over oxygen level. Our devices mimic the interface between the hypoxic
infarct zone and neighboring healthy tissue post-infarction and provide new insights into hypoxianormoxia crosstalk in cardiac cells. Other approaches discussed in Chapter 1 have similarly
modeled the dynamic cellular microenvironment of post-infarct myocardium and identified new
oxygen, stiffness, and strain-dependent mechanisms that underlie pathological remodeling of
many cardiac cell types. These technological advancements will help to identify new therapeutic
targets for myocardial infarction and can complement animal models of myocardial infarction to
expedite the drug discovery pipeline.
New model systems have sometimes produced conflicting results, which may be due to
the inherent heterogeneity of the biological responses or may highlight a need for more
standardized experimental methods. The development of models that recapitulate true infarcted
myocardium will also require more in vivo or ex vivo analyses to characterize injured myocardium
through techniques such as atomic force microscopy, fluorescent oxygen probes, and highresolution imaging. In addition, there are still many gaps in our understanding of cardiac cell
responses to spatial and temporal gradients beyond those described above, such as the
biochemical gradients that orchestrate the inflammatory cascade after myocardial infarction.
Recent approaches to model biochemical gradients using microfluidics301, 302 will pave the way for
more complex modeling of cytokine and chemokine gradients in the context of myocardial

97
infarction. Emerging fabrication methods that allow control over the spatiotemporal distribution of
biochemical factors in 3-D synthetic hydrogels303 could also be implemented to mimic the dynamic
nature of the extracellular matrix after infarction. These dynamic cues can also be combined on
chip for enhanced physiological relevance as a drug screening platform.
Future work can also focus on expanding models to engineer cardiac tissues with distinct
control over the positioning of multiple cell types and matrix components, leading to more granular
models of post-infarct myocardium. For example, model systems can incorporate relevant cell
types beyond cardiac myocytes and fibroblasts, such as neurons304 and immune cells305
. As the
work in this dissertation repeatedly implicated oxygen-dependent crosstalk with fibroblasts in proinflammatory signaling, future directions can dissect fibroblast-immune cell interactions. In
addition, as in vitro models have been predominantly 2-D monocultures that may be
micropatterned to control tissue architecture or 3-D cocultured tissues or spheroids for which
tissue architecture is relatively random, emerging methods to pattern multiple cell types in 2-D306
and 3-D307, 308 will improve the architectural relevance and reproducibility of engineered
multicellular tissues. 3-D bioprinting has also advanced considerably in recent years to provide
increasing structural complexity309, 310, including spatial gradients in porosity311 and material and
cell composition312, and can be implemented to make more precise tissue models.
Another critical consideration for in vitro models of post-infarct myocardium is cell source.
Existing models of post-infarct myocardium have mostly included primary cardiac cells from other
species, such as neonatal rats, because they have historically been the most accessible cardiac
cell source. However, cells from neonatal rats exhibit species-specific differences and are
relatively resistant to hypoxia, a key feature of the infarct microenvironment36. Thus, model
systems will be improved as the field moves towards human induced pluripotent stem cell
(hiPSC)-derived cardiac myocytes, which currently exhibit limited maturity in vitro. Recent
approaches to mature hiPSC-derived cardiac myocytes with electromechanical or biochemical
stimuli may help resolve this concern271, 313. In addition to providing human relevance, hiPSC-

98
derived cardiac myocytes can also enable the identification of genetic contributions to myocardial
infarction314 and development of patient-specific disease models and treatment regimens.
Lastly, it is critical to make these Infarct on a Chip systems more high-throughput and
scalable to have an impact on drug discovery. Future work should transition towards scalable
fabrication methods, such as rapid, multimaterial bioprinting of cardiac biowire scaffolds that are
compatible with 96-well plate formats315, and the development of substrates with integrated
electrodes to streamline electrical stimulation316
. In our work, while we were able to improve
fabrication throughput through 3D-printing microfluidic templates, it is important to note that
experimental throughput was hindered by our reliance on external gas cylinder infrastructure.
More scalable methods for controlling oxygen availability to cells have been developed and
described in Chapter 1, and these include harnessing gas-impermeable materials32, chemical
hypoxia reagents65
, and inherent diffusion limitations of spheroids84
. Throughput can also be
improved by integrating sensors for real-time readouts of parameters such as tissue contractility92,
93, action potentials317, the consumption or secretion of biomolecules318, 319, or physical aspects of
the microenvironment, such as oxygen concentration and temperature320. These types of multisensor systems will provide more continuous and detailed insight into cellular phenotypes in
response to drug treatments while also requiring less manual handling.
5.3. Final Conclusions
In summary, timely resolution of both inflammation and fibrosis are critical to minimizing
infarct size and the resulting extent of fibrotic remodeling. The work in this dissertation
demonstrated that oxygen-dependent crosstalk with human cardiac fibroblasts was repeatedly
implicated in cytokine signaling, resembling the initial inflammatory phase of wound healing postinfarction. Future work can focus on (1) elucidating the underlying mechanisms that drive this
exacerbation of inflammatory signaling after hypoxia-normoxia crosstalk or (2) leveraging this
therapeutically, for example through targeting fibroblast cytokine secretion in vivo, which has been

99
unexplored. This knowledge can likely extend beyond the context of cardiac fibrosis and into that
of other hypoxic diseases, such as cancer, in which hypoxic cancer-associated fibroblasts
similarly play a key role in disease progression321
.
From a technological perspective, engineered models of post-infarct myocardium have
exciting potential to address many of the gaps presented by oversimplified 2-D cell culture models
and animal models that lack human relevance. As technologies continue to develop, nextgeneration 4-dimensional (4-D) models could provide simultaneous control over spatial and
temporal changes in the physical, biochemical, and mechanical cues that correspond to the
phases of infarct healing. When further advanced with patient-derived cells, scalable fabrication
techniques, and integrated sensors, these 4-D models will emerge as new standards for disease
modeling and drug screening and lead to new breakthrough therapies for mitigating post-infarction
remodeling.

100
Resources
The work in this dissertation was funded by the NSF Graduate Research Fellow Program
Grant No. DGE-1842487, NIH grant R01 HL153286-01, and NIH grant K99/R00 HL157722. We
thank the USC Libraries Bioinformatics Service, and especially Yibu Chen, for providing guidance
in the analysis of transcriptomic data as well as access to the bioinformatics software and
computing resources. We also thank the Applied Genomics, Computation, and Translational
(AGCT) Core at Cedars-Sinai Medical Center for supporting the RNA-sequencing performed in
this work. HiPSCs were generously gifted by the laboratory of Dr. Justin Ichida at the Keck School
of Medicine.

101
References
(1) Cardiovascular diseases. World Health Organization, 2020. https://www.who.int/healthtopics/cardiovascular-diseases/#tab=tab_1 (accessed.
(2) Heart attack. National Heart, Lung, and Blood Institute, https://www.nhlbi.nih.gov/healthtopics/heart-attack (accessed.
(3) Torio, C. M.; Moore, B. J. National Inpatient Hospital Costs: The Most Expensive Conditions
by Payer. Agency for Healthcare Research and Quality, 2016. https://www.hcupus.ahrq.gov/reports/statbriefs/sb204-Most-Expensive-Hospital-Conditions.jsp (accessed.
(4) Heart Disease Facts. Centers for Disease Control and Prevention, 2020.
https://www.cdc.gov/heartdisease/facts.htm (accessed.
(5) Thygesen, K.; Alpert, J. S.; Jaffe, A. S.; Chaitman, B. R.; Bax, J. J.; Morrow, D. A.; White, H.
D.; for the Universal Definition of Myocardial Infarction, E. G. o. b. o. t. J. E. S. o. C. E. A. C. o. C.
A. A. H. A. A. W. H. F. W. T. F. Fourth Universal Definition of Myocardial Infarction (2018).
Circulation 2018, 138 (20), e618-e651. DOI: 10.1161/CIR.0000000000000617.
(6) Mythili, S.; Malathi, N. Diagnostic markers of acute myocardial infarction. Biomed Rep 2015,
3 (6), 743-748. DOI: 10.3892/br.2015.500.
(7) Bagai, A.; Dangas, G. D.; Stone, G. W.; Granger, C. B. Reperfusion strategies in acute
coronary syndromes. Circ Res 2014, 114 (12), 1918-1928. DOI:
10.1161/CIRCRESAHA.114.302744.
(8) Montalescot, G.; Andersen, H. R.; Antoniucci, D.; Betriu, A.; de Boer, M. J.; Grip, L.; Neumann,
F. J.; Rothman, M. T. Recommendations on percutaneous coronary intervention for the
reperfusion of acute ST elevation myocardial infarction. Heart 2004, 90 (6), e37.
(9) Myocardial Infarction: Management and Treatment. Cleveland Clinic, 2019.
https://my.clevelandclinic.org/health/diseases/16818-heart-attack-myocardialinfarction/management-and-treatment (accessed.
(10) Mercado, M. G.; Smith, D. K.; McConnon, M. L. Myocardial infarction: management of the
subacute period. Am Fam Physician 2013, 88 (9), 581-588.

102
(11) Iravanian, S.; Dudley, S. C. The renin-angiotensin-aldosterone system (RAAS) and cardiac
arrhythmias. Heart Rhythm 2008, 5 (6 Suppl), S12-17. DOI: 10.1016/j.hrthm.2008.02.025.
(12) Talman, V.; Ruskoaho, H. Cardiac fibrosis in myocardial infarction-from repair and
remodeling to regeneration. Cell Tissue Res 2016, 365 (3), 563-581. DOI: 10.1007/s00441-016-
2431-9.
(13) Aldosterone antagonists. Heart and Stroke Foundation of Canada,
https://www.heartandstroke.ca/heart/treatments/medications/aldosterone-antagonists
(accessed.
(14) StatPearls. 2024.
(15) Yadid, M.; Lind, J. U.; Ardoña, H. A. M.; Sheehy, S. P.; Dickinson, L. E.; Eweje, F.; Bastings,
M. M. C.; Pope, B.; O'Connor, B. B.; Straubhaar, J. R.; et al. Endothelial extracellular vesicles
contain protective proteins and rescue ischemia-reperfusion injury in a human heart-on-chip. Sci
Transl Med 2020, 12 (565). DOI: 10.1126/scitranslmed.aax8005.
(16) Cohn, J. N.; Ferrari, R.; Sharpe, N. Cardiac remodeling--concepts and clinical implications:
a consensus paper from an international forum on cardiac remodeling. Behalf of an International
Forum on Cardiac Remodeling. J Am Coll Cardiol 2000, 35 (3), 569-582. DOI: 10.1016/s0735-
1097(99)00630-0.
(17) Lorell, B. H.; Carabello, B. A. Left ventricular hypertrophy: pathogenesis, detection, and
prognosis. Circulation 2000, 102 (4), 470-479. DOI: 10.1161/01.cir.102.4.470.
(18) Konstam, M. A.; Kramer, D. G.; Patel, A. R.; Maron, M. S.; Udelson, J. E. Left ventricular
remodeling in heart failure: current concepts in clinical significance and assessment. JACC
Cardiovasc Imaging 2011, 4 (1), 98-108. DOI: 10.1016/j.jcmg.2010.10.008.
(19) Pfeffer, M. A.; Braunwald, E. Ventricular remodeling after myocardial infarction. Experimental
observations and clinical implications. Circulation 1990, 81 (4), 1161-1172. DOI:
10.1161/01.cir.81.4.1161.
(20) Azevedo, P. S.; Polegato, B. F.; Minicucci, M. F.; Paiva, S. A.; Zornoff, L. A. Cardiac
Remodeling: Concepts, Clinical Impact, Pathophysiological Mechanisms and Pharmacologic
Treatment. Arq Bras Cardiol 2016, 106 (1), 62-69. DOI: 10.5935/abc.20160005.
(21) Korup, E.; Dalsgaard, D.; Nyvad, O.; Jensen, T. M.; Toft, E.; Berning, J. Comparison of
degrees of left ventricular dilation within three hours and up to six days after onset of first acute

103
myocardial infarction. Am J Cardiol 1997, 80 (4), 449-453. DOI: 10.1016/s0002-9149(97)00393-
7.
(22) Sutton, M. G.; Sharpe, N. Left Ventricular Remodeling After Myocardial Infarction:
Pathophysiology and Therapy. Circulation, 2000; Vol. 101, pp 2981-2988.
(23) Rumberger, J. A.; Behrenbeck, T.; Breen, J. R.; Reed, J. E.; Gersh, B. J. Nonparallel changes
in global left ventricular chamber volume and muscle mass during the first year after transmural
myocardial infarction in humans. J Am Coll Cardiol 1993, 21 (3), 673-682. DOI: 10.1016/0735-
1097(93)90100-f.
(24) Armstrong, P. W. Left ventricular dysfunction: causes, natural history, and hopes for reversal.
Heart 2000, 84 Suppl 1, i15-17:discussion i50. DOI: 10.1136/heart.84.suppl_1.i15.
(25) Chew, D. S.; Heikki, H.; Schmidt, G.; Kavanagh, K. M.; Dommasch, M.; Bloch Thomsen, P.
E.; Sinnecker, D.; Raatikainen, P.; Exner, D. V. Change in Left Ventricular Ejection Fraction
Following First Myocardial Infarction and Outcome. JACC Clin Electrophysiol 2018, 4 (5), 672-
682. DOI: 10.1016/j.jacep.2017.12.015.
(26) Morrow, D. A. Myocardial Infarction: A Companion to Braunwald's Heart Disease. Elsevier:
St. Louis, Missouri, 2017; pp 449-450.
(27) Weber, K. T.; Sun, Y.; Bhattacharya, S. K.; Ahokas, R. A.; Gerling, I. C. Myofibroblastmediated mechanisms of pathological remodelling of the heart. Nat Rev Cardiol 2013, 10 (1), 15-
26. DOI: 10.1038/nrcardio.2012.158.
(28) Prabhu, S. D.; Frangogiannis, N. G. The Biological Basis for Cardiac Repair After Myocardial
Infarction: From Inflammation to Fibrosis. Circ Res 2016, 119 (1), 91-112. DOI:
10.1161/CIRCRESAHA.116.303577.
(29) Frangogiannis, N. G. The inflammatory response in myocardial injury, repair, and
remodelling. Nat Rev Cardiol 2014, 11 (5), 255-265. DOI: 10.1038/nrcardio.2014.28.
(30) Shinde, A. V.; Frangogiannis, N. G. Fibroblasts in myocardial infarction: a role in inflammation
and repair. J Mol Cell Cardiol 2014, 70, 74-82. DOI: 10.1016/j.yjmcc.2013.11.015.
(31) Rog-Zielinska, E. A.; Norris, R. A.; Kohl, P.; Markwald, R. The Living Scar--Cardiac
Fibroblasts and the Injured Heart. Trends Mol Med 2016, 22 (2), 99-114. DOI:
10.1016/j.molmed.2015.12.006.

104
(32) Mosadegh, B.; Dabiri, B. E.; Lockett, M. R.; Derda, R.; Campbell, P.; Parker, K. K.;
Whitesides, G. M. Three-dimensional paper-based model for cardiac ischemia. Adv Healthc Mater
2014, 3 (7), 1036-1043. DOI: 10.1002/adhm.201300575.
(33) Chiu, L. L.; Radisic, M.; Vunjak-Novakovic, G. Bioactive scaffolds for engineering
vascularized cardiac tissues. Macromol Biosci 2010, 10 (11), 1286-1301. DOI:
10.1002/mabi.201000202.
(34) Sen, C. K.; Khanna, S.; Roy, S. Perceived hyperoxia: oxygen-induced remodeling of the
reoxygenated heart. Cardiovasc Res 2006, 71 (2), 280-288. DOI:
10.1016/j.cardiores.2006.01.003.
(35) Rumsey, W. L.; Pawlowski, M.; Lejavardi, N.; Wilson, D. F. Oxygen pressure distribution in
the heart in vivo and evaluation of the ischemic "border zone". Am J Physiol 1994, 266 (4 Pt 2),
H1676-1680. DOI: 10.1152/ajpheart.1994.266.4.H1676.
(36) Lindsey, M. L.; Bolli, R.; Canty, J. M.; Du, X. J.; Frangogiannis, N. G.; Frantz, S.; Gourdie, R.
G.; Holmes, J. W.; Jones, S. P.; Kloner, R. A.; et al. Guidelines for experimental models of
myocardial ischemia and infarction. Am J Physiol Heart Circ Physiol 2018, 314 (4), H812-H838.
DOI: 10.1152/ajpheart.00335.2017.
(37) Muñoz-Sánchez, J.; Chánez-Cárdenas, M. E. The use of cobalt chloride as a chemical
hypoxia model. J Appl Toxicol 2019, 39 (4), 556-570. DOI: 10.1002/jat.3749.
(38) Zhao, R. Z.; Jiang, S.; Ru, N. Y.; Jiao, B.; Yu, Z. B. Comparison of hypoxic effects induced
by chemical and physical hypoxia on cardiomyocytes. Can J Physiol Pharmacol 2019, 97 (10),
980-988. DOI: 10.1139/cjpp-2019-0092.
(39) Pavlacky, J.; Polak, J. Technical Feasibility and Physiological Relevance of Hypoxic Cell
Culture Models. Front Endocrinol (Lausanne) 2020, 11, 57. DOI: 10.3389/fendo.2020.00057.
(40) Danon, A.; Zeevi-Levin, N.; Pinkovich, D. Y.; Michaeli, T.; Berkovich, A.; Flugelman, M.;
Eldar, Y. C.; Rosen, M. R.; Binah, O. Hypoxia causes connexin 43 internalization in neonatal rat
ventricular myocytes. Gen Physiol Biophys 2010, 29 (3), 222-233. DOI:
10.4149/gpb_2010_03_222.
(41) Zeevi-Levin, N.; Barac, Y. D.; Reisner, Y.; Reiter, I.; Yaniv, G.; Meiry, G.; Abassi, Z.; Kostin,
S.; Schaper, J.; Rosen, M. R.; et al. Gap junctional remodeling by hypoxia in cultured neonatal
rat ventricular myocytes. Cardiovasc Res 2005, 66 (1), 64-73. DOI:
10.1016/j.cardiores.2005.01.014.

105
(42) Turner, M. S.; Haywood, G. A.; Andreka, P.; You, L.; Martin, P. E.; Evans, W. H.; Webster,
K. A.; Bishopric, N. H. Reversible connexin 43 dephosphorylation during hypoxia and
reoxygenation is linked to cellular ATP levels. Circ Res 2004, 95 (7), 726-733. DOI:
10.1161/01.RES.0000144805.11519.1e.
(43) Xu, K. Y.; Kuppusamy, S. P.; Wang, J. Q.; Li, H.; Cui, H.; Dawson, T. M.; Huang, P. L.;
Burnett, A. L.; Kuppusamy, P.; Becker, L. C. Nitric oxide protects cardiac sarcolemmal membrane
enzyme function and ion active transport against ischemia-induced inactivation. J Biol Chem
2003, 278 (43), 41798-41803. DOI: 10.1074/jbc.M306865200.
(44) Yakushev, S.; Band, M.; Tissot van Patot, M. C.; Gassmann, M.; Avivi, A.; Bogdanova, A.
Cross talk between S-nitrosylation and S-glutathionylation in control of the Na,K-ATPase
regulation in hypoxic heart. Am J Physiol Heart Circ Physiol 2012, 303 (11), H1332-1343. DOI:
10.1152/ajpheart.00145.2012.
(45) González-Rodríguez, P.; Falcón, D.; Castro, M. J.; Ureña, J.; López-Barneo, J.; Castellano,
A. Hypoxic induction of T-type Ca(2+) channels in rat cardiac myocytes: role of HIF-1α and
RhoA/ROCK signalling. J Physiol 2015, 593 (21), 4729-4745. DOI: 10.1113/JP271053.
(46) Chen, S. J.; Bradley, M. E.; Lee, T. C. Chemical hypoxia triggers apoptosis of cultured
neonatal rat cardiac myocytes: modulation by calcium-regulated proteases and protein kinases.
Mol Cell Biochem 1998, 178 (1-2), 141-149. DOI: 10.1023/a:1006893528428.
(47) Tanaka, M.; Ito, H.; Adachi, S.; Akimoto, H.; Nishikawa, T.; Kasajima, T.; Marumo, F.; Hiroe,
M. Hypoxia induces apoptosis with enhanced expression of Fas antigen messenger RNA in
cultured neonatal rat cardiomyocytes. Circ Res 1994, 75 (3), 426-433. DOI:
10.1161/01.res.75.3.426.
(48) Long, X.; Boluyt, M. O.; Hipolito, M. L.; Lundberg, M. S.; Zheng, J. S.; O'Neill, L.; Cirielli, C.;
Lakatta, E. G.; Crow, M. T. p53 and the hypoxia-induced apoptosis of cultured neonatal rat cardiac
myocytes. J Clin Invest 1997, 99 (11), 2635-2643. DOI: 10.1172/JCI119452.
(49) Kang, P. M.; Haunstetter, A.; Aoki, H.; Usheva, A.; Izumo, S. Morphological and molecular
characterization of adult cardiomyocyte apoptosis during hypoxia and reoxygenation. Circ Res
2000, 87 (2), 118-125. DOI: 10.1161/01.res.87.2.118.
(50) Gao, Y.; Chu, M.; Hong, J.; Shang, J.; Xu, D. Hypoxia induces cardiac fibroblast proliferation
and phenotypic switch: a role for caveolae and caveolin-1/PTEN mediated pathway. J Thorac Dis
2014, 6 (10), 1458-1468. DOI: 10.3978/j.issn.2072-1439.2014.08.31.

106
(51) Watson, C. J.; Collier, P.; Tea, I.; Neary, R.; Watson, J. A.; Robinson, C.; Phelan, D.;
Ledwidge, M. T.; McDonald, K. M.; McCann, A.; et al. Hypoxia-induced epigenetic modifications
are associated with cardiac tissue fibrosis and the development of a myofibroblast-like phenotype.
Hum Mol Genet 2014, 23 (8), 2176-2188. DOI: 10.1093/hmg/ddt614.
(52) Zhao, X.; Wang, K.; Liao, Y.; Zeng, Q.; Li, Y.; Hu, F.; Liu, Y.; Meng, K.; Qian, C.; Zhang, Q.;
et al. MicroRNA-101a inhibits cardiac fibrosis induced by hypoxia via targeting TGFβRI on cardiac
fibroblasts. Cell Physiol Biochem 2015, 35 (1), 213-226. DOI: 10.1159/000369689.
(53) Tamamori, M.; Ito, H.; Hiroe, M.; Marumo, F.; Hata, R. I. Stimulation of collagen synthesis in
rat cardiac fibroblasts by exposure to hypoxic culture conditions and suppression of the effect by
natriuretic peptides. Cell Biol Int 1997, 21 (3), 175-180. DOI: 10.1006/cbir.1997.0130.
(54) Chu, W.; Li, X.; Li, C.; Wan, L.; Shi, H.; Song, X.; Liu, X.; Chen, X.; Zhang, C.; Shan, H.; et
al. TGFBR3, a potential negative regulator of TGF-β signaling, protects cardiac fibroblasts from
hypoxia-induced apoptosis. J Cell Physiol 2011, 226 (10), 2586-2594. DOI: 10.1002/jcp.22604.
(55) Zhao, X.; Wang, K.; Hu, F.; Qian, C.; Guan, H.; Feng, K.; Zhou, Y.; Chen, Z. MicroRNA-101
protects cardiac fibroblasts from hypoxia-induced apoptosis via inhibition of the TGF-β signaling
pathway. Int J Biochem Cell Biol 2015, 65, 155-164. DOI: 10.1016/j.biocel.2015.06.005.
(56) Yang, B.; He, K.; Zheng, F.; Wan, L.; Yu, X.; Wang, X.; Zhao, D.; Bai, Y.; Chu, W.; Sun, Y.;
et al. Over-expression of hypoxia-inducible factor-1 alpha in vitro protects the cardiac fibroblasts
from hypoxia-induced apoptosis. J Cardiovasc Med (Hagerstown) 2014, 15 (7), 579-586. DOI:
10.2459/JCM.0b013e3283629c52.
(57) Ladoux, A.; Frelin, C. Hypoxia is a strong inducer of vascular endothelial growth factor mRNA
expression in the heart. Biochem Biophys Res Commun 1993, 195 (2), 1005-1010. DOI:
10.1006/bbrc.1993.2144.
(58) Hwang, J. M.; Weng, Y. J.; Lin, J. A.; Bau, D. T.; Ko, F. Y.; Tsai, F. J.; Tsai, C. H.; Wu, C. H.;
Lin, P. C.; Huang, C. Y.; et al. Hypoxia-induced compensatory effect as related to Shh and HIF1alpha in ischemia embryo rat heart. Mol Cell Biochem 2008, 311 (1-2), 179-187. DOI:
10.1007/s11010-008-9708-6.
(59) Ontoria-Oviedo, I.; Dorronsoro, A.; Sánchez, R.; Ciria, M.; Gómez-Ferrer, M.; Buigues, M.;
Grueso, E.; Tejedor, S.; García-García, F.; González-King, H.; et al. Extracellular Vesicles
Secreted by Hypoxic AC10 Cardiomyocytes Modulate Fibroblast Cell Motility. Front Cardiovasc
Med 2018, 5, 152. DOI: 10.3389/fcvm.2018.00152.

107
(60) Shi, H.; Zhang, X.; He, Z.; Wu, Z.; Rao, L.; Li, Y. Metabolites of Hypoxic Cardiomyocytes
Induce the Migration of Cardiac Fibroblasts. Cell Physiol Biochem 2017, 41 (1), 413-421. DOI:
10.1159/000456531.
(61) Shivakumar, K.; Sollott, S. J.; Sangeetha, M.; Sapna, S.; Ziman, B.; Wang, S.; Lakatta, E. G.
Paracrine effects of hypoxic fibroblast-derived factors on the MPT-ROS threshold and viability of
adult rat cardiac myocytes. Am J Physiol Heart Circ Physiol 2008, 294 (6), H2653-2658. DOI:
10.1152/ajpheart.91443.2007.
(62) Gao, Q.; Guo, M.; Zeng, W.; Wang, Y.; Yang, L.; Pang, X.; Li, H.; Suo, Y.; Jiang, X.; Yu, C.
Matrix metalloproteinase 9 secreted by hypoxia cardiac fibroblasts triggers cardiac stem cell
migration in vitro. Stem Cells Int 2015, 2015, 836390. DOI: 10.1155/2015/836390.
(63) Wang, J. H.; Zhao, L.; Pan, X.; Chen, N. N.; Chen, J.; Gong, Q. L.; Su, F.; Yan, J.; Zhang,
Y.; Zhang, S. H. Hypoxia-stimulated cardiac fibroblast production of IL-6 promotes myocardial
fibrosis via the TGF-β1 signaling pathway. Lab Invest 2016, 96 (9), 1035. DOI:
10.1038/labinvest.2016.84.
(64) Cosme, J.; Guo, H.; Hadipour-Lakmehsari, S.; Emili, A.; Gramolini, A. O. Hypoxia-Induced
Changes in the Fibroblast Secretome, Exosome, and Whole-Cell Proteome Using Cultured,
Cardiac-Derived Cells Isolated from Neonatal Mice. J Proteome Res 2017, 16 (8), 2836-2847.
DOI: 10.1021/acs.jproteome.7b00144.
(65) Ren, L.; Liu, W.; Wang, Y.; Wang, J. C.; Tu, Q.; Xu, J.; Liu, R.; Shen, S. F.; Wang, J.
Investigation of hypoxia-induced myocardial injury dynamics in a tissue interface mimicking
microfluidic device. Anal Chem 2013, 85 (1), 235-244. DOI: 10.1021/ac3025812.
(66) Kang, Y. B. A.; Eo, J.; Bulutoglu, B.; Yarmush, M. L.; Usta, O. B. Progressive hypoxia-on-achip: An in vitro oxygen gradient model for capturing the effects of hypoxia on primary hepatocytes
in health and disease. Biotechnol Bioeng 2020, 117 (3), 763-775. DOI: 10.1002/bit.27225.
(67) Wang, L.; Liu, W.; Wang, Y.; Wang, J. C.; Tu, Q.; Liu, R.; Wang, J. Construction of oxygen
and chemical concentration gradients in a single microfluidic device for studying tumor cell-drug
interactions in a dynamic hypoxia microenvironment. Lab Chip 2013, 13 (4), 695-705. DOI:
10.1039/c2lc40661f.
(68) Chang, C. W.; Cheng, Y. J.; Tu, M.; Chen, Y. H.; Peng, C. C.; Liao, W. H.; Tung, Y. C. A
polydimethylsiloxane-polycarbonate hybrid microfluidic device capable of generating
perpendicular chemical and oxygen gradients for cell culture studies. Lab Chip 2014, 14 (19),
3762-3772. DOI: 10.1039/c4lc00732h.

108
(69) Shih, H. C.; Lee, T. A.; Wu, H. M.; Ko, P. L.; Liao, W. H.; Tung, Y. C. Microfluidic Collective
Cell Migration Assay for Study of Endothelial Cell Proliferation and Migration under Combinations
of Oxygen Gradients, Tensions, and Drug Treatments. Sci Rep 2019, 9 (1), 8234. DOI:
10.1038/s41598-019-44594-5.
(70) Rexius-Hall, M. L.; Rehman, J.; Eddington, D. T. A microfluidic oxygen gradient demonstrates
differential activation of the hypoxia-regulated transcription factors HIF-1α and HIF-2α. Integr Biol
(Camb) 2017, 9 (9), 742-750. DOI: 10.1039/c7ib00099e.
(71) Lo, J. F.; Sinkala, E.; Eddington, D. T. Oxygen gradients for open well cellular cultures via
microfluidic substrates. Lab Chip 2010, 10 (18), 2394-2401. DOI: 10.1039/c004660d.
(72) Polinkovsky, M.; Gutierrez, E.; Levchenko, A.; Groisman, A. Fine temporal control of the
medium gas content and acidity and on-chip generation of series of oxygen concentrations for
cell cultures. Lab Chip 2009, 9 (8), 1073-1084. DOI: 10.1039/b816191g.
(73) Lavrentieva, A. Advances in Biochemical Engineering/Biotechnology. Springer, 2020; pp 1-
25.
(74) Blatchley, M.; Park, K. M.; Gerecht, S. Designer Hydrogels for Precision Control of Oxygen
Tension and Mechanical Properties. J Mater Chem B 2015, 3 (40), 7939-7949. DOI:
10.1039/C5TB01038A.
(75) Park, K. M.; Gerecht, S. Hypoxia-inducible hydrogels. Nat Commun 2014, 5, 4075. DOI:
10.1038/ncomms5075.
(76) Boyce, M. W.; Simke, W. C.; Kenney, R. M.; Lockett, M. R. Generating linear oxygen
gradients across 3D cell cultures with block-layered oxygen controlled chips (BLOCCs). Anal
Methods 2020, 12 (1), 18-24. DOI: 10.1039/C9AY01690B.
(77) Pedron, S.; Becka, E.; Harley, B. A. Spatially gradated hydrogel platform as a 3D engineered
tumor microenvironment. Adv Mater 2015, 27 (9), 1567-1572. DOI: 10.1002/adma.201404896.
(78) Khademhosseini, A.; Eng, G.; Yeh, J.; Kucharczyk, P. A.; Langer, R.; Vunjak-Novakovic, G.;
Radisic, M. Microfluidic patterning for fabrication of contractile cardiac organoids. Biomed
Microdevices 2007, 9 (2), 149-157. DOI: 10.1007/s10544-006-9013-7.
(79) Chiu, L. L.; Iyer, R. K.; King, J. P.; Radisic, M. Biphasic electrical field stimulation aids in
tissue engineering of multicell-type cardiac organoids. Tissue Eng Part A 2011, 17 (11-12), 1465-
1477. DOI: 10.1089/ten.tea.2007.0244.

109
(80) Mills, R. J.; Titmarsh, D. M.; Koenig, X.; Parker, B. L.; Ryall, J. G.; Quaife-Ryan, G. A.; Voges,
H. K.; Hodson, M. P.; Ferguson, C.; Drowley, L.; et al. Functional screening in human cardiac
organoids reveals a metabolic mechanism for cardiomyocyte cell cycle arrest. Proc Natl Acad Sci
U S A 2017, 114 (40), E8372-E8381. DOI: 10.1073/pnas.1707316114.
(81) Iyer, R. K.; Odedra, D.; Chiu, L. L.; Vunjak-Novakovic, G.; Radisic, M. Vascular endothelial
growth factor secretion by nonmyocytes modulates Connexin-43 levels in cardiac organoids.
Tissue Eng Part A 2012, 18 (17-18), 1771-1783. DOI: 10.1089/ten.TEA.2011.0468.
(82) Iyer, R. K.; Chui, J.; Radisic, M. Spatiotemporal tracking of cells in tissue-engineered cardiac
organoids. J Tissue Eng Regen Med 2009, 3 (3), 196-207. DOI: 10.1002/term.153.
(83) Cai, Y.; Zhang, J.; Wu, J.; Li, Z. Y. Oxygen transport in a three-dimensional microvascular
network incorporated with early tumour growth and preexisting vessel cooption: numerical
simulation study. Biomed Res Int 2015, 2015, 476964. DOI: 10.1155/2015/476964.
(84) Richards, D. J.; Li, Y.; Kerr, C. M.; Yao, J.; Beeson, G. C.; Coyle, R. C.; Chen, X.; Jia, J.;
Damon, B.; Wilson, R.; et al. Human cardiac organoids for the modelling of myocardial infarction
and drug cardiotoxicity. Nat Biomed Eng 2020, 4 (4), 446-462. DOI: 10.1038/s41551-020-0539-
4.
(85) Liu, H.; Bolonduro, O. A.; Hu, N.; Ju, J.; Rao, A. A.; Duffy, B. M.; Huang, Z.; Black, L. D.;
Timko, B. P. Heart-on-a-Chip Model with Integrated Extra- and Intracellular Bioelectronics for
Monitoring Cardiac Electrophysiology under Acute Hypoxia. Nano Lett 2020, 20 (4), 2585-2593.
DOI: 10.1021/acs.nanolett.0c00076.
(86) Martewicz, S.; Michielin, F.; Serena, E.; Zambon, A.; Mongillo, M.; Elvassore, N. Reversible
alteration of calcium dynamics in cardiomyocytes during acute hypoxia transient in a microfluidic
platform. Integr Biol (Camb) 2012, 4 (2), 153-164. DOI: 10.1039/c1ib00087j.
(87) Bursac, N.; Parker, K. K.; Iravanian, S.; Tung, L. Cardiomyocyte cultures with controlled
macroscopic anisotropy: a model for functional electrophysiological studies of cardiac muscle.
Circ Res 2002, 91 (12), e45-54. DOI: 10.1161/01.res.0000047530.88338.eb.
(88) Feinberg, A. W.; Alford, P. W.; Jin, H.; Ripplinger, C. M.; Werdich, A. A.; Sheehy, S. P.;
Grosberg, A.; Parker, K. K. Controlling the contractile strength of engineered cardiac muscle by
hierarchal tissue architecture. Biomaterials 2012, 33 (23), 5732-5741. DOI:
10.1016/j.biomaterials.2012.04.043.
(89) Petersen, A. P.; Lyra-Leite, D. M.; Ariyasinghe, N. R.; Cho, N.; Goodwin, C. M.; Kim, J. Y.;
McCain, M. L. Microenvironmental Modulation of Calcium Wave Propagation Velocity in

110
Engineered Cardiac Tissues. Cell Mol Bioeng 2018, 11 (5), 337-352. DOI: 10.1007/s12195-018-
0522-2.
(90) McCain, M. L.; Yuan, H.; Pasqualini, F. S.; Campbell, P. H.; Parker, K. K. Matrix elasticity
regulates the optimal cardiac myocyte shape for contractility. Am J Physiol Heart Circ Physiol
2014, 306 (11), H1525-1539. DOI: 10.1152/ajpheart.00799.2013.
(91) Watson, J. L.; Aich, S.; Oller-Salvia, B.; Drabek, A. A.; Blacklow, S. C.; Chin, J.; Derivery, E.
High-efficacy subcellular micropatterning of proteins using fibrinogen anchors. J Cell Biol 2021,
220 (2). DOI: 10.1083/jcb.202009063.
(92) Lind, J. U.; Yadid, M.; Perkins, I.; O'Connor, B. B.; Eweje, F.; Chantre, C. O.; Hemphill, M.
A.; Yuan, H.; Campbell, P. H.; Vlassak, J. J.; et al. Cardiac microphysiological devices with flexible
thin-film sensors for higher-throughput drug screening. Lab Chip 2017, 17 (21), 3692-3703. DOI:
10.1039/c7lc00740j.
(93) Lind, J. U.; Busbee, T. A.; Valentine, A. D.; Pasqualini, F. S.; Yuan, H.; Yadid, M.; Park, S.
J.; Kotikian, A.; Nesmith, A. P.; Campbell, P. H.; et al. Instrumented cardiac microphysiological
devices via multimaterial three-dimensional printing. Nat Mater 2017, 16 (3), 303-308. DOI:
10.1038/nmat4782.
(94) Herum, K. M.; Lunde, I. G.; McCulloch, A. D.; Christensen, G. The Soft- and HardHeartedness of Cardiac Fibroblasts: Mechanotransduction Signaling Pathways in Fibrosis of the
Heart. J Clin Med 2017, 6 (5). DOI: 10.3390/jcm6050053.
(95) van Putten, S.; Shafieyan, Y.; Hinz, B. Mechanical control of cardiac myofibroblasts. J Mol
Cell Cardiol 2016, 93, 133-142. DOI: 10.1016/j.yjmcc.2015.11.025.
(96) Berry, M. F.; Engler, A. J.; Woo, Y. J.; Pirolli, T. J.; Bish, L. T.; Jayasankar, V.; Morine, K. J.;
Gardner, T. J.; Discher, D. E.; Sweeney, H. L. Mesenchymal stem cell injection after myocardial
infarction improves myocardial compliance. Am J Physiol Heart Circ Physiol 2006, 290 (6),
H2196-2203. DOI: 10.1152/ajpheart.01017.2005.
(97) Litwin, S. E.; Litwin, C. M.; Raya, T. E.; Warner, A. L.; Goldman, S. Contractility and stiffness
of noninfarcted myocardium after coronary ligation in rats. Effects of chronic angiotensin
converting enzyme inhibition. Circulation 1991, 83 (3), 1028-1037. DOI: 10.1161/01.cir.83.3.1028.
(98) Rusu, M.; Hilse, K.; Schuh, A.; Martin, L.; Slabu, I.; Stoppe, C.; Liehn, E. A. Biomechanical
assessment of remote and postinfarction scar remodeling following myocardial infarction. Sci Rep
2019, 9 (1), 16744. DOI: 10.1038/s41598-019-53351-7.

111
(99) Torres, W. M.; Jacobs, J.; Doviak, H.; Barlow, S. C.; Zile, M. R.; Shazly, T.; Spinale, F. G.
Regional and temporal changes in left ventricular strain and stiffness in a porcine model of
myocardial infarction. Am J Physiol Heart Circ Physiol 2018, 315 (4), H958-H967. DOI:
10.1152/ajpheart.00279.2018.
(100) Landry, N. M.; Rattan, S. G.; Dixon, I. M. C. An Improved Method of Maintaining Primary
Murine Cardiac Fibroblasts in Two-Dimensional Cell Culture. Sci Rep 2019, 9 (1), 12889. DOI:
10.1038/s41598-019-49285-9.
(101) Ariyasinghe, N. R.; Lyra-Leite, D. M.; McCain, M. L. Engineering cardiac microphysiological
systems to model pathological extracellular matrix remodeling. Am J Physiol Heart Circ Physiol
2018, 315 (4), H771-H789. DOI: 10.1152/ajpheart.00110.2018.
(102) Hu, W.; Wang, Z.; Xiao, Y.; Zhang, S.; Wang, J. Advances in crosslinking strategies of
biomedical hydrogels. Biomater Sci 2019, 7 (3), 843-855. DOI: 10.1039/c8bm01246f.
(103) Zhu, J.; Marchant, R. E. Design properties of hydrogel tissue-engineering scaffolds. Expert
Rev Med Devices 2011, 8 (5), 607-626. DOI: 10.1586/erd.11.27.
(104) Lyra-Leite, D. M.; Andres, A. M.; Petersen, A. P.; Ariyasinghe, N. R.; Cho, N.; Lee, J. A.;
Gottlieb, R. A.; McCain, M. L. Mitochondrial function in engineered cardiac tissues is regulated by
extracellular matrix elasticity and tissue alignment. Am J Physiol Heart Circ Physiol 2017, 313 (4),
H757-H767. DOI: 10.1152/ajpheart.00290.2017.
(105) Lyra-Leite, D. M.; Petersen, A. P.; Ariyasinghe, N. R.; Cho, N.; McCain, M. L. Mitochondrial
architecture in cardiac myocytes depends on cell shape and matrix rigidity. J Mol Cell Cardiol
2021, 150, 32-43. DOI: 10.1016/j.yjmcc.2020.10.004.
(106) Lyra-Leite, D. M.; Andres, A. M.; Cho, N.; Petersen, A. P.; Ariyasinghe, N. R.; Kim, S. S.;
Gottlieb, R. A.; McCain, M. L. Matrix-guided control of mitochondrial function in cardiac myocytes.
Acta Biomater 2019, 97, 281-295. DOI: 10.1016/j.actbio.2019.08.007.
(107) Guo, J.; Simmons, D. W.; Ramahdita, G.; Munsell, M. K.; Oguntuyo, K.; Kandalaft, B.; Rios,
B.; Pear, M.; Schuftan, D.; Jiang, H.; et al. Elastomer-Grafted iPSC-Derived Micro Heart Muscles
to Investigate Effects of Mechanical Loading on Physiology. ACS Biomaterials Science &
Engineering 2020. DOI: 10.1021/acsbiomaterials.0c00318.
(108) Jacot, J. G.; McCulloch, A. D.; Omens, J. H. Substrate stiffness affects the functional
maturation of neonatal rat ventricular myocytes. Biophys J 2008, 95 (7), 3479-3487. DOI:
10.1529/biophysj.107.124545.

112
(109) Boothe, S. D.; Myers, J. D.; Pok, S.; Sun, J.; Xi, Y.; Nieto, R. M.; Cheng, J.; Jacot, J. G. The
Effect of Substrate Stiffness on Cardiomyocyte Action Potentials. Cell Biochem Biophys 2016, 74
(4), 527-535. DOI: 10.1007/s12013-016-0758-1.
(110) Engler, A. J.; Carag-Krieger, C.; Johnson, C. P.; Raab, M.; Tang, H. Y.; Speicher, D. W.;
Sanger, J. W.; Sanger, J. M.; Discher, D. E. Embryonic cardiomyocytes beat best on a matrix with
heart-like elasticity: scar-like rigidity inhibits beating. J Cell Sci 2008, 121 (Pt 22), 3794-3802. DOI:
10.1242/jcs.029678.
(111) Bhana, B.; Iyer, R. K.; Chen, W. L.; Zhao, R.; Sider, K. L.; Likhitpanichkul, M.; Simmons, C.
A.; Radisic, M. Influence of substrate stiffness on the phenotype of heart cells. Biotechnol Bioeng
2010, 105 (6), 1148-1160. DOI: 10.1002/bit.22647.
(112) Pasqualini, F. S.; Agarwal, A.; O'Connor, B. B.; Liu, Q.; Sheehy, S. P.; Parker, K. K. Traction
force microscopy of engineered cardiac tissues. PLoS One 2018, 13 (3), e0194706. DOI:
10.1371/journal.pone.0194706.
(113) Bajaj, P.; Tang, X.; Saif, T. A.; Bashir, R. Stiffness of the substrate influences the phenotype
of embryonic chicken cardiac myocytes. J Biomed Mater Res A 2010, 95 (4), 1261-1269. DOI:
10.1002/jbm.a.32951.
(114) Hersch, N.; Wolters, B.; Dreissen, G.; Springer, R.; Kirchgeßner, N.; Merkel, R.; Hoffmann,
B. The constant beat: cardiomyocytes adapt their forces by equal contraction upon environmental
stiffening. Biol Open 2013, 2 (3), 351-361. DOI: 10.1242/bio.20133830.
(115) McCain, M. L.; Lee, H.; Aratyn-Schaus, Y.; Kléber, A. G.; Parker, K. K. Cooperative coupling
of cell-matrix and cell-cell adhesions in cardiac muscle. Proc Natl Acad Sci U S A 2012, 109 (25),
9881-9886. DOI: 10.1073/pnas.1203007109.
(116) Ariyasinghe, N. R.; Reck, C. H.; Viscio, A. A.; Petersen, A. P.; Lyra-Leite, D. M.; Cho, N.;
McCain, M. L. Engineering micromyocardium to delineate cellular and extracellular regulation of
myocardial tissue contractility. Integr Biol (Camb) 2017, 9 (9), 730-741. DOI: 10.1039/c7ib00081b.
(117) Nguyen, D. T.; Nagarajan, N.; Zorlutuna, P. Effect of Substrate Stiffness on Mechanical
Coupling and Force Propagation at the Infarct Boundary. Biophys J 2018, 115 (10), 1966-1980.
DOI: 10.1016/j.bpj.2018.08.050.
(118) Cho, N.; Razipour, S. E.; McCain, M. L. Featured Article: TGF-β1 dominates extracellular
matrix rigidity for inducing differentiation of human cardiac fibroblasts to myofibroblasts. Exp Biol
Med (Maywood) 2018, 243 (7), 601-612. DOI: 10.1177/1535370218761628.

113
(119) Goffin, J. M.; Pittet, P.; Csucs, G.; Lussi, J. W.; Meister, J. J.; Hinz, B. Focal adhesion size
controls tension-dependent recruitment of alpha-smooth muscle actin to stress fibers. J Cell Biol
2006, 172 (2), 259-268. DOI: 10.1083/jcb.200506179.
(120) Liu, F.; Lagares, D.; Choi, K. M.; Stopfer, L.; Marinković, A.; Vrbanac, V.; Probst, C. K.;
Hiemer, S. E.; Sisson, T. H.; Horowitz, J. C.; et al. Mechanosignaling through YAP and TAZ drives
fibroblast activation and fibrosis. Am J Physiol Lung Cell Mol Physiol 2015, 308 (4), L344-357.
DOI: 10.1152/ajplung.00300.2014.
(121) Herum, K. M.; Choppe, J.; Kumar, A.; Engler, A. J.; McCulloch, A. D. Mechanical regulation
of cardiac fibroblast profibrotic phenotypes. Mol Biol Cell 2017, 28 (14), 1871-1882. DOI:
10.1091/mbc.E17-01-0014.
(122) Marinković, A.; Mih, J. D.; Park, J. A.; Liu, F.; Tschumperlin, D. J. Improved throughput
traction microscopy reveals pivotal role for matrix stiffness in fibroblast contractility and TGF-β
responsiveness. Am J Physiol Lung Cell Mol Physiol 2012, 303 (3), L169-180. DOI:
10.1152/ajplung.00108.2012.
(123) Shen, J.; Xie, Y.; Liu, Z.; Zhang, S.; Wang, Y.; Jia, L.; Cai, Z.; Ma, H.; Xiang, M. Increased
myocardial stiffness activates cardiac microvascular endothelial cell via VEGF paracrine signaling
in cardiac hypertrophy. J Mol Cell Cardiol 2018, 122, 140-151. DOI: 10.1016/j.yjmcc.2018.08.014.
(124) Ceccato, T. L.; Starbuck, R. B.; Hall, J. K.; Walker, C. J.; Brown, T. E.; Killgore, J. P.; Anseth,
K. S.; Leinwand, L. A. Defining the Cardiac Fibroblast Secretome in a Fibrotic Microenvironment.
J Am Heart Assoc 2020, 9 (19), e017025. DOI: 10.1161/JAHA.120.017025.
(125) Li, C.; Ouyang, L.; Armstrong, J. P. K.; Stevens, M. M. Advances in the Fabrication of
Biomaterials for Gradient Tissue Engineering. Trends Biotechnol 2021, 39 (2), 150-164. DOI:
10.1016/j.tibtech.2020.06.005.
(126) Tse, J. R.; Engler, A. J. Stiffness gradients mimicking in vivo tissue variation regulate
mesenchymal stem cell fate. PLoS One 2011, 6 (1), e15978. DOI: 10.1371/journal.pone.0015978.
(127) Major, L. G.; Holle, A. W.; Young, J. L.; Hepburn, M. S.; Jeong, K.; Chin, I. L.; Sanderson,
R. W.; Jeong, J. H.; Aman, Z. M.; Kennedy, B. F.; et al. Volume Adaptation Controls Stem Cell
Mechanotransduction. ACS Appl Mater Interfaces 2019, 11 (49), 45520-45530. DOI:
10.1021/acsami.9b19770.
(128) Marklein, R. A.; Burdick, J. A. Spatially controlled hydrogel mechanics to modulate stem
cell interactions. 2010; Vol. 6.

114
(129) Kim, C.; Young, J. L.; Holle, A. W.; Jeong, K.; Major, L. G.; Jeong, J. H.; Aman, Z. M.; Han,
D. W.; Hwang, Y.; Spatz, J. P.; et al. Stem Cell Mechanosensation on Gelatin Methacryloyl
(GelMA) Stiffness Gradient Hydrogels. Ann Biomed Eng 2020, 48 (2), 893-902. DOI:
10.1007/s10439-019-02428-5.
(130) Sunyer, R.; Jin, A. J.; Nossal, R.; Sackett, D. L. Fabrication of hydrogels with steep stiffness
gradients for studying cell mechanical response. PLoS One 2012, 7 (10), e46107. DOI:
10.1371/journal.pone.0046107.
(131) Dou, J.; Mao, S.; Li, H.; Lin, J. M. Combination Stiffness Gradient with Chemical Stimulation
Directs Glioma Cell Migration on a Microfluidic Chip. Anal Chem 2020, 92 (1), 892-898. DOI:
10.1021/acs.analchem.9b03681.
(132) Hadden, W. J.; Young, J. L.; Holle, A. W.; McFetridge, M. L.; Kim, D. Y.; Wijesinghe, P.;
Taylor-Weiner, H.; Wen, J. H.; Lee, A. R.; Bieback, K.; et al. Stem cell migration and
mechanotransduction on linear stiffness gradient hydrogels. Proc Natl Acad Sci U S A 2017, 114
(22), 5647-5652. DOI: 10.1073/pnas.1618239114.
(133) Wang, P. Y.; Tsai, W. B.; Voelcker, N. H. Screening of rat mesenchymal stem cell behaviour
on polydimethylsiloxane stiffness gradients. Acta Biomater 2012, 8 (2), 519-530. DOI:
10.1016/j.actbio.2011.09.030.
(134) Lavrentieva, A.; Fleischhammer, T.; Enders, A.; Pirmahboub, H.; Bahnemann, J.;
Pepelanova, I. Fabrication of Stiffness Gradients of GelMA Hydrogels Using a 3D Printed
Micromixer. Macromol Biosci 2020, 20 (7), e2000107. DOI: 10.1002/mabi.202000107.
(135) Zhao, H.; Li, X.; Zhao, S.; Zeng, Y.; Zhao, L.; Ding, H.; Sun, W.; Du, Y. Microengineered in
vitro model of cardiac fibrosis through modulating myofibroblast mechanotransduction.
Biofabrication 2014, 6 (4), 045009. DOI: 10.1088/1758-5082/6/4/045009.
(136) Liu, F.; Mih, J. D.; Shea, B. S.; Kho, A. T.; Sharif, A. S.; Tager, A. M.; Tschumperlin, D. J.
Feedback amplification of fibrosis through matrix stiffening and COX-2 suppression. J Cell Biol
2010, 190 (4), 693-706. DOI: 10.1083/jcb.201004082.
(137) Kloxin, A. M.; Kasko, A. M.; Salinas, C. N.; Anseth, K. S. Photodegradable hydrogels for
dynamic tuning of physical and chemical properties. Science 2009, 324 (5923), 59-63. DOI:
10.1126/science.1169494.
(138) Crocini, C.; Walker, C. J.; Anseth, K. S.; Leinwand, L. A. Three-dimensional encapsulation
of adult mouse cardiomyocytes in hydrogels with tunable stiffness. Prog Biophys Mol Biol 2020,
154, 71-79. DOI: 10.1016/j.pbiomolbio.2019.04.008.

115
(139) Young, J. L.; Engler, A. J. Hydrogels with time-dependent material properties enhance
cardiomyocyte differentiation in vitro. Biomaterials 2011, 32 (4), 1002-1009. DOI:
10.1016/j.biomaterials.2010.10.020.
(140) Burdick, J. A.; Murphy, W. L. Moving from static to dynamic complexity in hydrogel design.
Nat Commun 2012, 3, 1269. DOI: 10.1038/ncomms2271.
(141) Galie, P. A.; Westfall, M. V.; Stegemann, J. P. Reduced serum content and increased matrix
stiffness promote the cardiac myofibroblast transition in 3D collagen matrices. Cardiovasc Pathol
2011, 20 (6), 325-333. DOI: 10.1016/j.carpath.2010.10.001.
(142) Yu, J.; Seldin, M. M.; Fu, K.; Li, S.; Lam, L.; Wang, P.; Wang, Y.; Huang, D.; Nguyen, T. L.;
Wei, B.; et al. Topological Arrangement of Cardiac Fibroblasts Regulates Cellular Plasticity. Circ
Res 2018, 123 (1), 73-85. DOI: 10.1161/CIRCRESAHA.118.312589.
(143) van Spreeuwel, A. C. C.; Bax, N. A. M.; van Nierop, B. J.; Aartsma-Rus, A.; Goumans, M.
T. H.; Bouten, C. V. C. Mimicking Cardiac Fibrosis in a Dish: Fibroblast Density Rather than
Collagen Density Weakens Cardiomyocyte Function. J Cardiovasc Transl Res 2017, 10 (2), 116-
127. DOI: 10.1007/s12265-017-9737-1.
(144) van Spreeuwel, A. C.; Bax, N. A.; Bastiaens, A. J.; Foolen, J.; Loerakker, S.; Borochin, M.;
van der Schaft, D. W.; Chen, C. S.; Baaijens, F. P.; Bouten, C. V. The influence of matrix
(an)isotropy on cardiomyocyte contraction in engineered cardiac microtissues. Integr Biol (Camb)
2014, 6 (4), 422-429. DOI: 10.1039/c3ib40219c.
(145) DePalma, S. J.; Davidson, C. D.; Stis, A. E.; Helms, A. S.; Baker, B. M. Microenvironmental
determinants of organized iPSC-cardiomyocyte tissues on synthetic fibrous matrices. Biomater
Sci 2021, 9 (1), 93-107. DOI: 10.1039/d0bm01247e.
(146) Ma, Z.; Huebsch, N.; Koo, S.; Mandegar, M. A.; Siemons, B.; Boggess, S.; Conklin, B. R.;
Grigoropoulos, C. P.; Healy, K. E. Contractile deficits in engineered cardiac microtissues as a
result of MYBPC3 deficiency and mechanical overload. Nat Biomed Eng 2018, 2 (12), 955-967.
DOI: 10.1038/s41551-018-0280-4.
(147) Baker, B. M.; Trappmann, B.; Wang, W. Y.; Sakar, M. S.; Kim, I. L.; Shenoy, V. B.; Burdick,
J. A.; Chen, C. S. Cell-mediated fibre recruitment drives extracellular matrix mechanosensing
in engineered fibrillar microenvironments. Nat Mater 2015, 14 (12), 1262-1268. DOI:
10.1038/nmat4444.

116
(148) Davidson, C. D.; Jayco, D. K. P.; Matera, D. L.; DePalma, S. J.; Hiraki, H. L.; Wang, W. Y.;
Baker, B. M. Myofibroblast activation in synthetic fibrous matrices composed of dextran vinyl
sulfone. Acta Biomater 2020, 105, 78-86. DOI: 10.1016/j.actbio.2020.01.009.
(149) Matera, D. L.; DiLillo, K. M.; Smith, M. R.; Davidson, C. D.; Parikh, R.; Said, M.; Wilke, C.
A.; Lombaert, I. M.; Arnold, K. B.; Moore, B. B.; et al. Microengineered 3D pulmonary interstitial
mimetics highlight a critical role for matrix degradation in myofibroblast differentiation. Sci Adv
2020, 6 (37). DOI: 10.1126/sciadv.abb5069.
(150) Wang, E. Y.; Rafatian, N.; Zhao, Y.; Lee, A.; Lai, B. F. L.; Lu, R. X.; Jekic, D.; Davenport
Huyer, L.; Knee-Walden, E. J.; Bhattacharya, S.; et al. Biowire Model of Interstitial and Focal
Cardiac Fibrosis. ACS Cent Sci 2019, 5 (7), 1146-1158. DOI: 10.1021/acscentsci.9b00052.
(151) Wang, H.; Haeger, S. M.; Kloxin, A. M.; Leinwand, L. A.; Anseth, K. S. Redirecting valvular
myofibroblasts into dormant fibroblasts through light-mediated reduction in substrate modulus.
PLoS One 2012, 7 (7), e39969. DOI: 10.1371/journal.pone.0039969.
(152) Gorcsan, J.; Tanaka, H. Echocardiographic assessment of myocardial strain. J Am Coll
Cardiol 2011, 58 (14), 1401-1413. DOI: 10.1016/j.jacc.2011.06.038.
(153) Fink, C.; Ergün, S.; Kralisch, D.; Remmers, U.; Weil, J.; Eschenhagen, T. Chronic stretch
of engineered heart tissue induces hypertrophy and functional improvement. FASEB J 2000, 14
(5), 669-679. DOI: 10.1096/fasebj.14.5.669.
(154) Zimmermann, W. H.; Schneiderbanger, K.; Schubert, P.; Didié, M.; Münzel, F.; Heubach, J.
F.; Kostin, S.; Neuhuber, W. L.; Eschenhagen, T. Tissue engineering of a differentiated cardiac
muscle construct. Circ Res 2002, 90 (2), 223-230. DOI: 10.1161/hh0202.103644.
(155) Salazar, B. H.; Cashion, A. T.; Dennis, R. G.; Birla, R. K. Development of a Cyclic Strain
Bioreactor for Mechanical Enhancement and Assessment of Bioengineered Myocardial
Constructs. Cardiovasc Eng Technol 2015, 6 (4), 533-545. DOI: 10.1007/s13239-015-0236-8.
(156) Massai, D.; Pisani, G.; Isu, G.; Rodriguez Ruiz, A.; Cerino, G.; Galluzzi, R.; Pisanu, A.;
Tonoli, A.; Bignardi, C.; Audenino, A. L.; et al. Bioreactor Platform for Biomimetic Culture and.
Front Bioeng Biotechnol 2020, 8, 733. DOI: 10.3389/fbioe.2020.00733.
(157) Tulloch, N. L.; Muskheli, V.; Razumova, M. V.; Korte, F. S.; Regnier, M.; Hauch, K. D.;
Pabon, L.; Reinecke, H.; Murry, C. E. Growth of engineered human myocardium with mechanical
loading and vascular coculture. Circ Res 2011, 109 (1), 47-59. DOI:
10.1161/CIRCRESAHA.110.237206.

117
(158) Naito, H.; Melnychenko, I.; Didié, M.; Schneiderbanger, K.; Schubert, P.; Rosenkranz, S.;
Eschenhagen, T.; Zimmermann, W. H. Optimizing engineered heart tissue for therapeutic
applications as surrogate heart muscle. Circulation 2006, 114 (1 Suppl), I72-78. DOI:
10.1161/CIRCULATIONAHA.105.001560.
(159) McCain, M. L.; Sheehy, S. P.; Grosberg, A.; Goss, J. A.; Parker, K. K. Recapitulating
maladaptive, multiscale remodeling of failing myocardium on a chip. Proc Natl Acad Sci U S A
2013, 110 (24), 9770-9775. DOI: 10.1073/pnas.1304913110.
(160) Occhetta, P.; Isu, G.; Lemme, M.; Conficconi, C.; Oertle, P.; Räz, C.; Visone, R.; Cerino,
G.; Plodinec, M.; Rasponi, M.; et al. A three-dimensional in vitro dynamic micro-tissue model of
cardiac scar formation. Integr Biol (Camb) 2018, 10 (3), 174-183. DOI: 10.1039/c7ib00199a.
(161) Balestrini, J. L.; Billiar, K. L. Equibiaxial cyclic stretch stimulates fibroblasts to rapidly
remodel fibrin. J Biomech 2006, 39 (16), 2983-2990. DOI: 10.1016/j.jbiomech.2005.10.025.
(162) Balestrini, J. L.; Billiar, K. L. Magnitude and duration of stretch modulate fibroblast
remodeling. J Biomech Eng 2009, 131 (5), 051005. DOI: 10.1115/1.3049527.
(163) Yokoyama, T.; Sekiguchi, K.; Tanaka, T.; Tomaru, K.; Arai, M.; Suzuki, T.; Nagai, R.
Angiotensin II and mechanical stretch induce production of tumor necrosis factor in cardiac
fibroblasts. Am J Physiol 1999, 276 (6), H1968-1976. DOI: 10.1152/ajpheart.1999.276.6.H1968.
(164) Wang, J.; Chen, H.; Seth, A.; McCulloch, C. A. Mechanical force regulation of myofibroblast
differentiation in cardiac fibroblasts. Am J Physiol Heart Circ Physiol 2003, 285 (5), H1871-1881.
DOI: 10.1152/ajpheart.00387.2003.
(165) Ugolini, G. S.; Rasponi, M.; Pavesi, A.; Santoro, R.; Kamm, R.; Fiore, G. B.; Pesce, M.;
Soncini, M. On-chip assessment of human primary cardiac fibroblasts proliferative responses to
uniaxial cyclic mechanical strain. Biotechnol Bioeng 2016, 113 (4), 859-869. DOI:
10.1002/bit.25847.
(166) Atance, J.; Yost, M. J.; Carver, W. Influence of the extracellular matrix on the regulation of
cardiac fibroblast behavior by mechanical stretch. J Cell Physiol 2004, 200 (3), 377-386. DOI:
10.1002/jcp.20034.
(167) Watson, C. J.; Phelan, D.; Collier, P.; Horgan, S.; Glezeva, N.; Cooke, G.; Xu, M.; Ledwidge,
M.; McDonald, K.; Baugh, J. A. Extracellular matrix sub-types and mechanical stretch impact
human cardiac fibroblast responses to transforming growth factor beta. Connect Tissue Res 2014,
55 (3), 248-256. DOI: 10.3109/03008207.2014.904856.

118
(168) Butt, R. P.; Bishop, J. E. Mechanical load enhances the stimulatory effect of serum growth
factors on cardiac fibroblast procollagen synthesis. J Mol Cell Cardiol 1997, 29 (4), 1141-1151.
DOI: 10.1006/jmcc.1996.0347.
(169) Li, X.; Garcia-Elias, A.; Benito, B.; Nattel, S. The effects of cardiac stretch on atrial
fibroblasts: Analysis of the evidence and potential role in atrial fibrillation. Cardiovasc Res 2021.
DOI: 10.1093/cvr/cvab035.
(170) Ugolini, G. S.; Pavesi, A.; Rasponi, M.; Fiore, G. B.; Kamm, R.; Soncini, M. Human cardiac
fibroblasts adaptive responses to controlled combined mechanical strain and oxygen changes in
vitro. Elife 2017, 6. DOI: 10.7554/eLife.22847.
(171) Rysä, J.; Tokola, H.; Ruskoaho, H. Mechanical stretch induced transcriptomic profiles in
cardiac myocytes. Sci Rep 2018, 8 (1), 4733. DOI: 10.1038/s41598-018-23042-w.
(172) Wang, B. W.; Wu, G. J.; Cheng, W. P.; Shyu, K. G. Mechanical stretch via transforming
growth factor-β1 activates microRNA-208a to regulate hypertrophy in cultured rat cardiac
myocytes. J Formos Med Assoc 2013, 112 (10), 635-643. DOI: 10.1016/j.jfma.2013.01.002.
(173) Liao, X. D.; Wang, X. H.; Jin, H. J.; Chen, L. Y.; Chen, Q. Mechanical stretch induces
mitochondria-dependent apoptosis in neonatal rat cardiomyocytes and G2/M accumulation in
cardiac fibroblasts. Cell Res 2004, 14 (1), 16-26. DOI: 10.1038/sj.cr.7290198.
(174) Law, M. R.; Watt, H. C.; Wald, N. J. The underlying risk of death after myocardial infarction
in the absence of treatment. Arch Intern Med 2002, 162 (21), 2405-2410. DOI:
10.1001/archinte.162.21.2405.
(175) Jenča, D.; Melenovský, V.; Stehlik, J.; Staněk, V.; Kettner, J.; Kautzner, J.; Adámková, V.;
Wohlfahrt, P. Heart failure after myocardial infarction: incidence and predictors. ESC Heart Fail
2021, 8 (1), 222-237. DOI: 10.1002/ehf2.13144.
(176) Nallamothu, B. K.; Bradley, E. H.; Krumholz, H. M. Time to treatment in primary
percutaneous coronary intervention. N Engl J Med 2007, 357 (16), 1631-1638. DOI:
10.1056/NEJMra065985.
(177) Banerjee, I.; Fuseler, J. W.; Price, R. L.; Borg, T. K.; Baudino, T. A. Determination of cell
types and numbers during cardiac development in the neonatal and adult rat and mouse. Am J
Physiol Heart Circ Physiol 2007, 293 (3), H1883-1891. DOI: 10.1152/ajpheart.00514.2007.

119
(178) Vliegen, H. W.; van der Laarse, A.; Cornelisse, C. J.; Eulderink, F. Myocardial changes in
pressure overload-induced left ventricular hypertrophy. A study on tissue composition,
polyploidization and multinucleation. Eur Heart J 1991, 12 (4), 488-494. DOI:
10.1093/oxfordjournals.eurheartj.a059928.
(179) Hinz, B.; Celetta, G.; Tomasek, J. J.; Gabbiani, G.; Chaponnier, C. Alpha-smooth muscle
actin expression upregulates fibroblast contractile activity. Mol Biol Cell 2001, 12 (9), 2730-2741.
DOI: 10.1091/mbc.12.9.2730.
(180) Shinde, A. V.; Humeres, C.; Frangogiannis, N. G. The role of α-smooth muscle actin in
fibroblast-mediated matrix contraction and remodeling. Biochim Biophys Acta Mol Basis Dis 2017,
1863 (1), 298-309. DOI: 10.1016/j.bbadis.2016.11.006.
(181) Janbandhu, V.; Tallapragada, V.; Patrick, R.; Li, Y.; Abeygunawardena, D.; Humphreys, D.
T.; Martin, E. M. M. A.; Ward, A. O.; Contreras, O.; Farbehi, N.; et al. Hif-1a suppresses ROSinduced proliferation of cardiac fibroblasts following myocardial infarction. Cell Stem Cell 2022,
29 (2), 281-297.e212. DOI: 10.1016/j.stem.2021.10.009.
(182) Livak, K. J.; Schmittgen, T. D. Analysis of relative gene expression data using real-time
quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 2001, 25 (4), 402-408. DOI:
10.1006/meth.2001.1262.
(183) Demichev, V.; Messner, C. B.; Vernardis, S. I.; Lilley, K. S.; Ralser, M. DIA-NN: neural
networks and interference correction enable deep proteome coverage in high throughput. Nat
Methods 2020, 17 (1), 41-44. DOI: 10.1038/s41592-019-0638-x.
(184) Tyanova, S.; Temu, T.; Sinitcyn, P.; Carlson, A.; Hein, M. Y.; Geiger, T.; Mann, M.; Cox, J.
The Perseus computational platform for comprehensive analysis of (prote)omics data. Nat
Methods 2016, 13 (9), 731-740. DOI: 10.1038/nmeth.3901.
(185) Semenza, G. L. Hypoxia-inducible factors in physiology and medicine. Cell 2012, 148 (3),
399-408. DOI: 10.1016/j.cell.2012.01.021.
(186) Koh, M. Y.; Powis, G. Passing the baton: the HIF switch. Trends Biochem Sci 2012, 37 (9),
364-372. DOI: 10.1016/j.tibs.2012.06.004.
(187) Hubbi, M. E.; Semenza, G. L. Regulation of cell proliferation by hypoxia-inducible factors.
Am J Physiol Cell Physiol 2015, 309 (12), C775-782. DOI: 10.1152/ajpcell.00279.2015.

120
(188) Siegel, P. M.; Massagué, J. Cytostatic and apoptotic actions of TGF-beta in homeostasis
and cancer. Nat Rev Cancer 2003, 3 (11), 807-821. DOI: 10.1038/nrc1208.
(189) McAnulty, R. J. Fibroblasts and myofibroblasts: their source, function and role in disease.
Int J Biochem Cell Biol 2007, 39 (4), 666-671. DOI: 10.1016/j.biocel.2006.11.005.
(190) Zeisberg, E. M.; Kalluri, R. Origins of cardiac fibroblasts. Circ Res 2010, 107 (11), 1304-
1312. DOI: 10.1161/CIRCRESAHA.110.231910.
(191) Forsythe, J. A.; Jiang, B. H.; Iyer, N. V.; Agani, F.; Leung, S. W.; Koos, R. D.; Semenza, G.
L. Activation of vascular endothelial growth factor gene transcription by hypoxia-inducible factor
1. Mol Cell Biol 1996, 16 (9), 4604-4613. DOI: 10.1128/MCB.16.9.4604.
(192) Pez, F.; Dayan, F.; Durivault, J.; Kaniewski, B.; Aimond, G.; Le Provost, G. S.; Deux, B.;
Clézardin, P.; Sommer, P.; Pouysségur, J.; et al. The HIF-1-inducible lysyl oxidase activates HIF1 via the Akt pathway in a positive regulation loop and synergizes with HIF-1 in promoting tumor
cell growth. Cancer Res 2011, 71 (5), 1647-1657. DOI: 10.1158/0008-5472.CAN-10-1516.
(193) Mia, M. M.; Cibi, D. M.; Ghani, S. A. B. A.; Singh, A.; Tee, N.; Sivakumar, V.; Bogireddi, H.;
Cook, S. A.; Mao, J.; Singh, M. K. Loss of Yap/Taz in cardiac fibroblasts attenuates adverse
remodelling and improves cardiac function. Cardiovasc Res 2022, 118 (7), 1785-1804. DOI:
10.1093/cvr/cvab205.
(194) Francisco, J.; Zhang, Y.; Jeong, J. I.; Mizushima, W.; Ikeda, S.; Ivessa, A.; Oka, S.; Zhai,
P.; Tallquist, M. D.; Del Re, D. P. Blockade of Fibroblast YAP Attenuates Cardiac Fibrosis and
Dysfunction Through MRTF-A Inhibition. JACC Basic Transl Sci 2020, 5 (9), 931-945. DOI:
10.1016/j.jacbts.2020.07.009.
(195) Merrick, W. C.; Pavitt, G. D. Protein Synthesis Initiation in Eukaryotic Cells. Cold Spring
Harb Perspect Biol 2018, 10 (12). DOI: 10.1101/cshperspect.a033092.
(196) Wu, B.; Rockel, J. S.; Lagares, D.; Kapoor, M. Ephrins and Eph Receptor Signaling in Tissue
Repair and Fibrosis. Curr Rheumatol Rep 2019, 21 (6), 23. DOI: 10.1007/s11926-019-0825-x.
(197) Zhu, X.; Zhang, X.; Gu, W.; Zhao, H.; Hao, S.; Ning, Z. ANGPTL4 suppresses the
profibrogenic functions of atrial fibroblasts induced by angiotensin II by up-regulating PPARγ. Iran
J Basic Med Sci 2023, 26 (5), 587-593. DOI: 10.22038/IJBMS.2023.69196.15077.
(198) Ali, A.; Shafarin, J.; Abu Jabal, R.; Aljabi, N.; Hamad, M.; Sualeh Muhammad, J.;
Unnikannan, H. Ferritin heavy chain (FTH1) exerts significant antigrowth effects in breast cancer

121
cells by inhibiting the expression of c-MYC. FEBS Open Bio 2021, 11 (11), 3101-3114. DOI:
10.1002/2211-5463.13303.
(199) Ellen, T. P.; Ke, Q.; Zhang, P.; Costa, M. NDRG1, a growth and cancer related gene:
regulation of gene expression and function in normal and disease states. Carcinogenesis 2008,
29 (1), 2-8. DOI: 10.1093/carcin/bgm200.
(200) Leask, A. Potential therapeutic targets for cardiac fibrosis: TGFbeta, angiotensin,
endothelin, CCN2, and PDGF, partners in fibroblast activation. Circ Res 2010, 106 (11), 1675-
1680. DOI: 10.1161/CIRCRESAHA.110.217737.
(201) Jin, L. The actin associated protein palladin in smooth muscle and in the development of
diseases of the cardiovasculature and in cancer. J Muscle Res Cell Motil 2011, 32 (1), 7-17. DOI:
10.1007/s10974-011-9246-9.
(202) Montégut, L.; Joseph, A.; Chen, H.; Abdellatif, M.; Ruckenstuhl, C.; Motiño, O.; Lambertucci,
F.; Anagnostopoulos, G.; Lachkar, S.; Dichtinger, S.; et al. High plasma concentrations of acylcoenzyme A binding protein (ACBP) predispose to cardiovascular disease: Evidence for a
phylogenetically conserved proaging function of ACBP. Aging Cell 2023, 22 (1), e13751. DOI:
10.1111/acel.13751.
(203) Spinelli, S.; Guida, L.; Vigliarolo, T.; Passalacqua, M.; Begani, G.; Magnone, M.; Sturla, L.;
Benzi, A.; Ameri, P.; Lazzarini, E.; et al. The ABA-LANCL1/2 Hormone-Receptors System
Protects H9c2 Cardiomyocytes from Hypoxia-Induced Mitochondrial Injury via an AMPK- and NOMediated Mechanism. Cells 2022, 11 (18). DOI: 10.3390/cells11182888.
(204) Zanetti, F.; Carpi, A.; Menabò, R.; Giorgio, M.; Schulz, R.; Valen, G.; Baysa, A.; Massimino,
M. L.; Sorgato, M. C.; Bertoli, A.; et al. The cellular prion protein counteracts cardiac oxidative
stress. Cardiovasc Res 2014, 104 (1), 93-102. DOI: 10.1093/cvr/cvu194.
(205) Sheu, J. J.; Chai, H. T.; Chiang, J. Y.; Sung, P. H.; Chen, Y. L.; Yip, H. K. Cellular Prion
Protein Is Essential for Myocardial Regeneration but Not the Recovery of Left Ventricular Function
from Apical Ballooning. Biomedicines 2022, 10 (1). DOI: 10.3390/biomedicines10010167.
(206) Brereton, C. J.; Yao, L.; Davies, E. R.; Zhou, Y.; Vukmirovic, M.; Bell, J. A.; Wang, S.; Ridley,
R. A.; Dean, L. S. N.; Andriotis, O. G.; et al. Pseudohypoxic HIF pathway activation dysregulates
collagen structure-function in human lung fibrosis. Elife 2022, 11. DOI: 10.7554/eLife.69348.
(207) Senavirathna, L. K.; Huang, C.; Pushparaj, S.; Xu, D.; Liu, L. Hypoxia and transforming
growth factor β1 regulation of long non-coding RNA transcriptomes in human pulmonary
fibroblasts. Physiol Rep 2020, 8 (1), e14343. DOI: 10.14814/phy2.14343.

122
(208) Wek, R. C. Role of eIF2α Kinases in Translational Control and Adaptation to Cellular Stress.
Cold Spring Harb Perspect Biol 2018, 10 (7). DOI: 10.1101/cshperspect.a032870.
(209) Murtha, L. A.; Schuliga, M. J.; Mabotuwana, N. S.; Hardy, S. A.; Waters, D. W.; Burgess, J.
K.; Knight, D. A.; Boyle, A. J. The Processes and Mechanisms of Cardiac and Pulmonary Fibrosis.
Front Physiol 2017, 8, 777. DOI: 10.3389/fphys.2017.00777.
(210) Hu, K.; Babapoor-Farrokhran, S.; Rodrigues, M.; Deshpande, M.; Puchner, B.;
Kashiwabuchi, F.; Hassan, S. J.; Asnaghi, L.; Handa, J. T.; Merbs, S.; et al. Hypoxia-inducible
factor 1 upregulation of both VEGF and ANGPTL4 is required to promote the angiogenic
phenotype in uveal melanoma. Oncotarget 2016, 7 (7), 7816-7828. DOI:
10.18632/oncotarget.6868.
(211) Daseke, M. J.; Tenkorang, M. A. A.; Chalise, U.; Konfrst, S. R.; Lindsey, M. L. Cardiac
fibroblast activation during myocardial infarction wound healing: Fibroblast polarization after MI.
Matrix Biol 2020, 91-92, 109-116. DOI: 10.1016/j.matbio.2020.03.010.
(212) Araki, K.; Morita, M.; Bederman, A. G.; Konieczny, B. T.; Kissick, H. T.; Sonenberg, N.;
Ahmed, R. Translation is actively regulated during the differentiation of CD8. Nat Immunol 2017,
18 (9), 1046-1057. DOI: 10.1038/ni.3795.
(213) Duś-Szachniewicz, K.; Gdesz-Birula, K.; Zduniak, K.; Wiśniewski, J. R. Proteomic-Based
Analysis of Hypoxia- and Physioxia-Responsive Proteins and Pathways in Diffuse Large B-Cell
Lymphoma. Cells 2021, 10 (8). DOI: 10.3390/cells10082025.
(214) Koumenis, C.; Naczki, C.; Koritzinsky, M.; Rastani, S.; Diehl, A.; Sonenberg, N.; Koromilas,
A.; Wouters, B. G. Regulation of protein synthesis by hypoxia via activation of the endoplasmic
reticulum kinase PERK and phosphorylation of the translation initiation factor eIF2alpha. Mol Cell
Biol 2002, 22 (21), 7405-7416. DOI: 10.1128/MCB.22.21.7405-7416.2002.
(215) Camelliti, P.; Green, C. R.; LeGrice, I.; Kohl, P. Fibroblast network in rabbit sinoatrial node:
structural and functional identification of homogeneous and heterogeneous cell coupling. Circ Res
2004, 94 (6), 828-835. DOI: 10.1161/01.RES.0000122382.19400.14.
(216) Vasquez, C.; Mohandas, P.; Louie, K. L.; Benamer, N.; Bapat, A. C.; Morley, G. E.
Enhanced fibroblast-myocyte interactions in response to cardiac injury. Circ Res 2010, 107 (8),
1011-1020. DOI: 10.1161/CIRCRESAHA.110.227421.
(217) Miragoli, M.; Gaudesius, G.; Rohr, S. Electrotonic modulation of cardiac impulse conduction
by myofibroblasts. Circ Res 2006, 98 (6), 801-810. DOI: 10.1161/01.RES.0000214537.44195.a3.

123
(218) He, K.; Shi, X.; Zhang, X.; Dang, S.; Ma, X.; Liu, F.; Xu, M.; Lv, Z.; Han, D.; Fang, X.; et al.
Long-distance intercellular connectivity between cardiomyocytes and cardiofibroblasts mediated
by membrane nanotubes. Cardiovasc Res 2011, 92 (1), 39-47. DOI: 10.1093/cvr/cvr189.
(219) Hall, C.; Gehmlich, K.; Denning, C.; Pavlovic, D. Complex Relationship Between Cardiac
Fibroblasts and Cardiomyocytes in Health and Disease. J Am Heart Assoc 2021, 10 (5), e019338.
DOI: 10.1161/JAHA.120.019338.
(220) Kluckova, K.; Tennant, D. A. Metabolic implications of hypoxia and pseudohypoxia in
pheochromocytoma and paraganglioma. Cell Tissue Res 2018, 372 (2), 367-378. DOI:
10.1007/s00441-018-2801-6.
(221) Mohlin, S.; Wigerup, C.; Jögi, A.; Påhlman, S. Hypoxia, pseudohypoxia and cellular
differentiation. Exp Cell Res 2017, 356 (2), 192-196. DOI: 10.1016/j.yexcr.2017.03.007.
(222) Liberti, M. V.; Locasale, J. W. The Warburg Effect: How Does it Benefit Cancer Cells?
Trends Biochem Sci 2016, 41 (3), 211-218. DOI: 10.1016/j.tibs.2015.12.001.
(223) Aquino-Gálvez, A.; González-Ávila, G.; Jiménez-Sánchez, L. L.; Maldonado-Martínez, H.
A.; Cisneros, J.; Toscano-Marquez, F.; Castillejos-López, M.; Torres-Espíndola, L. M.; VelázquezCruz, R.; Rodríguez, V. H. O.; et al. Dysregulated expression of hypoxia-inducible factors
augments myofibroblasts differentiation in idiopathic pulmonary fibrosis. Respir Res 2019, 20 (1),
130. DOI: 10.1186/s12931-019-1100-4.
(224) Serio, S.; Pagiatakis, C.; Musolino, E.; Felicetta, A.; Carullo, P.; Laura Frances, J.; Papa,
L.; Rozzi, G.; Salvarani, N.; Miragoli, M.; et al. Cardiac Aging Is Promoted by Pseudohypoxia
Increasing p300-Induced Glycolysis. Circ Res 2023, 133 (8), 687-703. DOI:
10.1161/CIRCRESAHA.123.322676.
(225) Laflamme, M. A.; Murry, C. E. Heart regeneration. Nature 2011, 473 (7347), 326-335. DOI:
10.1038/nature10147.
(226) Caulfield, J. B.; Leinbach, R.; Gold, H. The relationship of myocardial infarct size and
prognosis. Circulation 1976, 53 (3 Suppl), I141-144.
(227) Chen, B.; Frangogiannis, N. G. Chemokines in Myocardial Infarction. J Cardiovasc Transl
Res 2021, 14 (1), 35-52. DOI: 10.1007/s12265-020-10006-7.
(228) Dewald, O.; Zymek, P.; Winkelmann, K.; Koerting, A.; Ren, G.; Abou-Khamis, T.; Michael,
L. H.; Rollins, B. J.; Entman, M. L.; Frangogiannis, N. G. CCL2/Monocyte Chemoattractant

124
Protein-1 regulates inflammatory responses critical to healing myocardial infarcts. Circ Res 2005,
96 (8), 881-889. DOI: 10.1161/01.RES.0000163017.13772.3a.
(229) Bujak, M.; Dobaczewski, M.; Chatila, K.; Mendoza, L. H.; Li, N.; Reddy, A.; Frangogiannis,
N. G. Interleukin-1 receptor type I signaling critically regulates infarct healing and cardiac
remodeling. Am J Pathol 2008, 173 (1), 57-67. DOI: 10.2353/ajpath.2008.070974.
(230) Saxena, A.; Chen, W.; Su, Y.; Rai, V.; Uche, O. U.; Li, N.; Frangogiannis, N. G. IL-1 induces
proinflammatory leukocyte infiltration and regulates fibroblast phenotype in the infarcted
myocardium. J Immunol 2013, 191 (9), 4838-4848. DOI: 10.4049/jimmunol.1300725.
(231) Yan, X.; Anzai, A.; Katsumata, Y.; Matsuhashi, T.; Ito, K.; Endo, J.; Yamamoto, T.;
Takeshima, A.; Shinmura, K.; Shen, W.; et al. Temporal dynamics of cardiac immune cell
accumulation following acute myocardial infarction. J Mol Cell Cardiol 2013, 62, 24-35. DOI:
10.1016/j.yjmcc.2013.04.023.
(232) Khalil, N. N.; McCain, M. L. Engineering the Cellular Microenvironment of Post-infarct
Myocardium on a Chip. Front Cardiovasc Med 2021, 8, 709871. DOI: 10.3389/fcvm.2021.709871.
(233) Rexius-Hall, M. L.; Mauleon, G.; Malik, A. B.; Rehman, J.; Eddington, D. T. Microfluidic
platform generates oxygen landscapes for localized hypoxic activation. Lab Chip 2014, 14 (24),
4688-4695. DOI: 10.1039/c4lc01168f.
(234) Brennan, M. D.; Rexius-Hall, M. L.; Elgass, L. J.; Eddington, D. T. Oxygen control with
microfluidics. Lab Chip 2014, 14 (22), 4305-4318. DOI: 10.1039/c4lc00853g.
(235) Rexius-Hall, M. L.; Khalil, N. N.; Escopete, S. S.; Li, X.; Hu, J.; Yuan, H.; Parker, S. J.;
McCain, M. L. A myocardial infarct border-zone-on-a-chip demonstrates distinct regulation of
cardiac tissue function by an oxygen gradient. Sci Adv 2022, 8 (49), eabn7097. DOI:
10.1126/sciadv.abn7097.
(236) Khalil, N. N.; Petersen, A. P.; Song, C. J.; Chen, Y.; Takamoto, K.; Kellogg, A. C.; Chen, E.
Z.; McMahon, A. P.; McCain, M. L. User-friendly microfluidic system reveals native-like
morphological and transcriptomic phenotypes induced by shear stress in proximal tubule
epithelium. APL Bioeng 2023, 7 (3), 036106. DOI: 10.1063/5.0143614.
(237) Grist, S. M.; Chrostowski, L.; Cheung, K. C. Optical oxygen sensors for applications in
microfluidic cell culture. Sensors (Basel) 2010, 10 (10), 9286-9316. DOI: 10.3390/s101009286.

125
(238) Knutson, A. K.; Williams, A. L.; Boisvert, W. A.; Shohet, R. V. HIF in the heart: development,
metabolism, ischemia, and atherosclerosis. J Clin Invest 2021, 131 (17). DOI:
10.1172/JCI137557.
(239) Tanner, M. A.; Thomas, T. P.; Maitz, C. A.; Grisanti, L. A. β2-Adrenergic Receptors Increase
Cardiac Fibroblast Proliferation Through the Gαs/ERK1/2-Dependent Secretion of Interleukin-6.
Int J Mol Sci 2020, 21 (22). DOI: 10.3390/ijms21228507.
(240) Yap, J.; Irei, J.; Lozano-Gerona, J.; Vanapruks, S.; Bishop, T.; Boisvert, W. A. Macrophages
in cardiac remodelling after myocardial infarction. Nat Rev Cardiol 2023, 20 (6), 373-385. DOI:
10.1038/s41569-022-00823-5.
(241) Lee, G. W.; Lee, T. H.; Vilcek, J. TSG-14, a tumor necrosis factor- and IL-1-inducible protein,
is a novel member of the pentaxin family of acute phase proteins. J Immunol 1993, 150 (5), 1804-
1812.
(242) Garlanda, C.; Bottazzi, B.; Magrini, E.; Inforzato, A.; Mantovani, A. PTX3, a Humoral Pattern
Recognition Molecule, in Innate Immunity, Tissue Repair, and Cancer. Physiol Rev 2018, 98 (2),
623-639. DOI: 10.1152/physrev.00016.2017.
(243) Peri, G.; Introna, M.; Corradi, D.; Iacuitti, G.; Signorini, S.; Avanzini, F.; Pizzetti, F.;
Maggioni, A. P.; Moccetti, T.; Metra, M.; et al. PTX3, A prototypical long pentraxin, is an early
indicator of acute myocardial infarction in humans. Circulation 2000, 102 (6), 636-641. DOI:
10.1161/01.cir.102.6.636.
(244) Duran, S.; Duran, I.; Kaptanagası, F. A. O.; Nartop, F.; Ciftci, H.; Korkmaz, G. G. The role
of pentraxin 3 as diagnostic value in classification of patients with heart failure. Clin Biochem
2013, 46 (12), 983-987. DOI: 10.1016/j.clinbiochem.2013.04.026.
(245) Levy, J. A. The unexpected pleiotropic activities of RANTES. J Immunol 2009, 182 (7),
3945-3946. DOI: 10.4049/jimmunol.0990015.
(246) Ridker, P. M.; Rifai, N.; Pfeffer, M.; Sacks, F.; Lepage, S.; Braunwald, E. Elevation of tumor
necrosis factor-alpha and increased risk of recurrent coronary events after myocardial infarction.
Circulation 2000, 101 (18), 2149-2153. DOI: 10.1161/01.cir.101.18.2149.
(247) Mahmoud, A. H.; Taha, N. M.; Zakhary, M.; Tadros, M. S. PTEN gene & TNF-alpha in acute
myocardial infarction. Int J Cardiol Heart Vasc 2019, 23, 100366. DOI:
10.1016/j.ijcha.2019.100366.

126
(248) Venugopal, H.; Hanna, A.; Humeres, C.; Frangogiannis, N. G. Properties and Functions of
Fibroblasts and Myofibroblasts in Myocardial Infarction. Cells 2022, 11 (9). DOI:
10.3390/cells11091386.
(249) Bujak, M.; Frangogiannis, N. G. The role of TGF-beta signaling in myocardial infarction and
cardiac remodeling. Cardiovasc Res 2007, 74 (2), 184-195. DOI:
10.1016/j.cardiores.2006.10.002.
(250) Zou, J.; Fei, Q.; Xiao, H.; Wang, H.; Liu, K.; Liu, M.; Zhang, H.; Xiao, X.; Wang, K.; Wang,
N. VEGF-A promotes angiogenesis after acute myocardial infarction through increasing ROS
production and enhancing ER stress-mediated autophagy. J Cell Physiol 2019, 234 (10), 17690-
17703. DOI: 10.1002/jcp.28395.
(251) González-Santamaría, J.; Villalba, M.; Busnadiego, O.; López-Olañeta, M. M.; Sandoval,
P.; Snabel, J.; López-Cabrera, M.; Erler, J. T.; Hanemaaijer, R.; Lara-Pezzi, E.; et al. Matrix crosslinking lysyl oxidases are induced in response to myocardial infarction and promote cardiac
dysfunction. Cardiovasc Res 2016, 109 (1), 67-78. DOI: 10.1093/cvr/cvv214.
(252) Sandstedt, J.; Sandstedt, M.; Lundqvist, A.; Jansson, M.; Sopasakis, V. R.; Jeppsson, A.;
Hultén, L. M. Human cardiac fibroblasts isolated from patients with severe heart failure are
immune-competent cells mediating an inflammatory response. Cytokine 2019, 113, 319-325.
DOI: 10.1016/j.cyto.2018.09.021.
(253) Richardson, W. J.; Clarke, S. A.; Quinn, T. A.; Holmes, J. W. Physiological Implications of
Myocardial Scar Structure. Compr Physiol 2015, 5 (4), 1877-1909. DOI: 10.1002/cphy.c140067.
(254) Gulati, A.; Jabbour, A.; Ismail, T. F.; Guha, K.; Khwaja, J.; Raza, S.; Morarji, K.; Brown, T.
D.; Ismail, N. A.; Dweck, M. R.; et al. Association of fibrosis with mortality and sudden cardiac
death in patients with nonischemic dilated cardiomyopathy. JAMA 2013, 309 (9), 896-908. DOI:
10.1001/jama.2013.1363.
(255) Shi, Y.; Lin, S.; Staats, K. A.; Li, Y.; Chang, W. H.; Hung, S. T.; Hendricks, E.; Linares, G.
R.; Wang, Y.; Son, E. Y.; et al. Haploinsufficiency leads to neurodegeneration in C9ORF72
ALS/FTD human induced motor neurons. Nat Med 2018, 24 (3), 313-325. DOI: 10.1038/nm.4490.
(256) Lai, J. D.; Berlind, J. E.; Fricklas, G.; Lie, C.; Urenda, J. P.; Lam, K.; Sta Maria, N.; Jacobs,
R.; Yu, V.; Zhao, Z.; et al. KCNJ2 inhibition mitigates mechanical injury in a human brain organoid
model of traumatic brain injury. Cell Stem Cell 2024, 31 (4), 519-536.e518. DOI:
10.1016/j.stem.2024.03.004.

127
(257) Yang, X.; Rodriguez, M. L.; Leonard, A.; Sun, L.; Fischer, K. A.; Wang, Y.; Ritterhoff, J.;
Zhao, L.; Kolwicz, S. C.; Pabon, L.; et al. Fatty Acids Enhance the Maturation of Cardiomyocytes
Derived from Human Pluripotent Stem Cells. Stem Cell Reports 2019, 13 (4), 657-668. DOI:
10.1016/j.stemcr.2019.08.013.
(258) Kim, H. D. Expression of intermediate filament desmin and vimentin in the human fetal
heart. Anat Rec 1996, 246 (2), 271-278. DOI: 10.1002/(SICI)1097-0185(199610)246:2<271::AIDAR13>3.0.CO;2-L.
(259) Zuppinger, C.; Gibbons, G.; Dutta-Passecker, P.; Segiser, A.; Most, H.; Suter, T. M.
Characterization of cytoskeleton features and maturation status of cultured human iPSC-derived
cardiomyocytes. Eur J Histochem 2017, 61 (2), 2763. DOI: 10.4081/ejh.2017.2763.
(260) Mosaddeghzadeh, N.; Ahmadian, M. R. The RHO Family GTPases: Mechanisms of
Regulation and Signaling. Cells 2021, 10 (7). DOI: 10.3390/cells10071831.
(261) Ohtsuka, T.; Hamada, M.; Inoue, K.; Ohshima, K.; Sujzuki, J.; Matsunaka, T.; Ogimoto, A.;
Hara, Y.; Shigematsu, Y.; Higaki, J. Relation of circulating interleukin-6 to left ventricular
remodeling in patients with reperfused anterior myocardial infarction. Clin Cardiol 2004, 27 (7),
417-420. DOI: 10.1002/clc.4960270712.
(262) Kobara, M.; Noda, K.; Kitamura, M.; Okamoto, A.; Shiraishi, T.; Toba, H.; Matsubara, H.;
Nakata, T. Antibody against interleukin-6 receptor attenuates left ventricular remodelling after
myocardial infarction in mice. Cardiovasc Res 2010, 87 (3), 424-430. DOI: 10.1093/cvr/cvq078.
(263) Hilfiker-Kleiner, D.; Shukla, P.; Klein, G.; Schaefer, A.; Stapel, B.; Hoch, M.; Müller, W.;
Scherr, M.; Theilmeier, G.; Ernst, M.; et al. Continuous glycoprotein-130-mediated signal
transducer and activator of transcription-3 activation promotes inflammation, left ventricular
rupture, and adverse outcome in subacute myocardial infarction. Circulation 2010, 122 (2), 145-
155. DOI: 10.1161/CIRCULATIONAHA.109.933127.
(264) Wang, J. H.; Zhao, L.; Pan, X.; Chen, N. N.; Chen, J.; Gong, Q. L.; Su, F.; Yan, J.; Zhang,
Y.; Zhang, S. H. Hypoxia-stimulated cardiac fibroblast production of IL-6 promotes myocardial
fibrosis via the TGF-β1 signaling pathway. Lab Invest 2016, 96 (8), 839-852. DOI:
10.1038/labinvest.2016.65.
(265) Alter, C.; Henseler, A. S.; Owenier, C.; Hesse, J.; Ding, Z.; Lautwein, T.; Bahr, J.; Hayat,
S.; Kramann, R.; Kostenis, E.; et al. IL-6 in the infarcted heart is preferentially formed by
fibroblasts and modulated by purinergic signaling. J Clin Invest 2023, 133 (11). DOI:
10.1172/JCI163799.

128
(266) Wei, K.; Nguyen, H. N.; Brenner, M. B. Fibroblast pathology in inflammatory diseases. J
Clin Invest 2021, 131 (20). DOI: 10.1172/JCI149538.
(267) Zhang, J.; Tao, R.; Campbell, K. F.; Carvalho, J. L.; Ruiz, E. C.; Kim, G. C.; Schmuck, E.
G.; Raval, A. N.; da Rocha, A. M.; Herron, T. J.; et al. Functional cardiac fibroblasts derived from
human pluripotent stem cells via second heart field progenitors. Nat Commun 2019, 10 (1), 2238.
DOI: 10.1038/s41467-019-09831-5.
(268) Hidalgo, A.; Glass, N.; Ovchinnikov, D.; Yang, S. K.; Zhang, X.; Mazzone, S.; Chen, C.;
Wolvetang, E.; Cooper-White, J. Modelling ischemia-reperfusion injury (IRI). APL Bioeng 2018, 2
(2), 026102. DOI: 10.1063/1.5000746.
(269) Choi, S. C.; Seo, H. R.; Cui, L. H.; Song, M. H.; Noh, J. M.; Kim, K. S.; Choi, J. H.; Kim, J.
H.; Park, C. Y.; Joo, H. J.; et al. Modeling Hypoxic Stress In Vitro Using Human Embryonic Stem
Cells Derived Cardiomyocytes Matured by FGF4 and Ascorbic Acid Treatment. Cells 2021, 10
(10). DOI: 10.3390/cells10102741.
(270) Gentillon, C.; Li, D.; Duan, M.; Yu, W. M.; Preininger, M. K.; Jha, R.; Rampoldi, A.; Saraf,
A.; Gibson, G. C.; Qu, C. K.; et al. Targeting HIF-1α in combination with PPARα activation and
postnatal factors promotes the metabolic maturation of human induced pluripotent stem cellderived cardiomyocytes. J Mol Cell Cardiol 2019, 132, 120-135. DOI:
10.1016/j.yjmcc.2019.05.003.
(271) Ronaldson-Bouchard, K.; Ma, S. P.; Yeager, K.; Chen, T.; Song, L.; Sirabella, D.; Morikawa,
K.; Teles, D.; Yazawa, M.; Vunjak-Novakovic, G. Advanced maturation of human cardiac tissue
grown from pluripotent stem cells. Nature 2018, 556 (7700), 239-243. DOI: 10.1038/s41586-018-
0016-3.
(272) Zhao, Y.; Rafatian, N.; Feric, N. T.; Cox, B. J.; Aschar-Sobbi, R.; Wang, E. Y.; Aggarwal,
P.; Zhang, B.; Conant, G.; Ronaldson-Bouchard, K.; et al. A Platform for Generation of ChamberSpecific Cardiac Tissues and Disease Modeling. Cell 2019, 176 (4), 913-927.e918. DOI:
10.1016/j.cell.2018.11.042.
(273) Chen, T.; Vunjak-Novakovic, G. Human Tissue-Engineered Model of Myocardial IschemiaReperfusion Injury. Tissue Eng Part A 2019, 25 (9-10), 711-724. DOI:
10.1089/ten.TEA.2018.0212.
(274) Caruso, M.; Meurant, S.; Detraux, D.; Mathieu, A.; Gilson, M.; Dieu, M.; Fattaccioli, A.;
Demazy, C.; Najimi, M.; Sokal, E.; et al. The global downregulation of protein synthesis observed
during hepatogenic maturation is associated with a decrease in TOP mRNA translation. Stem Cell
Reports 2023, 18 (1), 254-268. DOI: 10.1016/j.stemcr.2022.11.020.

129
(275) Marcon, B. H.; Holetz, F. B.; Eastman, G.; Origa-Alves, A. C.; Amorós, M. A.; de Aguiar, A.
M.; Rebelatto, C. K.; Brofman, P. R. S.; Sotelo-Silveira, J.; Dallagiovanna, B. Downregulation of
the protein synthesis machinery is a major regulatory event during early adipogenic differentiation
of human adipose-derived stromal cells. Stem Cell Res 2017, 25, 191-201. DOI:
10.1016/j.scr.2017.10.027.
(276) Chung, E. S.; Packer, M.; Lo, K. H.; Fasanmade, A. A.; Willerson, J. T.; Investigators, A.-T.
T. A. C. H. F. Randomized, double-blind, placebo-controlled, pilot trial of infliximab, a chimeric
monoclonal antibody to tumor necrosis factor-alpha, in patients with moderate-to-severe heart
failure: results of the anti-TNF Therapy Against Congestive Heart Failure (ATTACH) trial.
Circulation 2003, 107 (25), 3133-3140. DOI: 10.1161/01.CIR.0000077913.60364.D2.
(277) Padfield, G. J.; Din, J. N.; Koushiappi, E.; Mills, N. L.; Robinson, S. D.; Cruden, N. e. M.;
Lucking, A. J.; Chia, S.; Harding, S. A.; Newby, D. E. Cardiovascular effects of tumour necrosis
factor α antagonism in patients with acute myocardial infarction: a first in human study. Heart
2013, 99 (18), 1330-1335. DOI: 10.1136/heartjnl-2013-303648.
(278) Rolski, F.; Tkacz, K.; Węglarczyk, K.; Kwiatkowski, G.; Pelczar, P.; Jaźwa-Kusior, A.; Bar,
A.; Kuster, G. M.; Chłopicki, S.; Siedlar, M.; et al. TNF-α protects from exacerbated myocarditis
and cardiac death by suppressing expansion of activated heart-reactive CD4+ T cells. Cardiovasc
Res 2024, 120 (1), 82-94. DOI: 10.1093/cvr/cvad158.
(279) Wu, J. J.; Poon, K. Y.; Channual, J. C.; Shen, A. Y. Association between tumor necrosis
factor inhibitor therapy and myocardial infarction risk in patients with psoriasis. Arch Dermatol
2012, 148 (11), 1244-1250. DOI: 10.1001/archdermatol.2012.2502.
(280) Low, A. S.; Symmons, D. P.; Lunt, M.; Mercer, L. K.; Gale, C. P.; Watson, K. D.; Dixon, W.
G.; Hyrich, K. L.; Consortium, B. S. f. R. B. R. f. R. A. B.-R. a. t. B. C. C. Relationship between
exposure to tumour necrosis factor inhibitor therapy and incidence and severity of myocardial
infarction in patients with rheumatoid arthritis. Ann Rheum Dis 2017, 76 (4), 654-660. DOI:
10.1136/annrheumdis-2016-209784.
(281) Broch, K.; Anstensrud, A. K.; Woxholt, S.; Sharma, K.; Tøllefsen, I. M.; Bendz, B.; Aakhus,
S.; Ueland, T.; Amundsen, B. H.; Damås, J. K.; et al. Randomized Trial of Interleukin-6 Receptor
Inhibition in Patients With Acute ST-Segment Elevation Myocardial Infarction. J Am Coll Cardiol
2021, 77 (15), 1845-1855. DOI: 10.1016/j.jacc.2021.02.049.
(282) Huse, C.; Anstensrud, A. K.; Michelsen, A. E.; Ueland, T.; Broch, K.; Woxholt, S.; Yang, K.;
Sharma, K.; Tøllefsen, I. M.; Bendz, B.; et al. Interleukin-6 inhibition in ST-elevation myocardial
infarction: Immune cell profile in the randomised ASSAIL-MI trial. EBioMedicine 2022, 80, 104013.
DOI: 10.1016/j.ebiom.2022.104013.

130
(283) Ridker, P. M.; Everett, B. M.; Thuren, T.; MacFadyen, J. G.; Chang, W. H.; Ballantyne, C.;
Fonseca, F.; Nicolau, J.; Koenig, W.; Anker, S. D.; et al. Antiinflammatory Therapy with
Canakinumab for Atherosclerotic Disease. N Engl J Med 2017, 377 (12), 1119-1131. DOI:
10.1056/NEJMoa1707914.
(284) Strand, V.; Ahadieh, S.; French, J.; Geier, J.; Krishnaswami, S.; Menon, S.; Checchio, T.;
Tensfeldt, T. G.; Hoffman, E.; Riese, R.; et al. Systematic review and meta-analysis of serious
infections with tofacitinib and biologic disease-modifying antirheumatic drug treatment in
rheumatoid arthritis clinical trials. Arthritis Res Ther 2015, 17, 362. DOI: 10.1186/s13075-015-
0880-2.
(285) Gittenberger-de Groot, A. C.; Vrancken Peeters, M. P.; Mentink, M. M.; Gourdie, R. G.;
Poelmann, R. E. Epicardium-derived cells contribute a novel population to the myocardial wall
and the atrioventricular cushions. Circ Res 1998, 82 (10), 1043-1052. DOI:
10.1161/01.res.82.10.1043.
(286) Moore-Morris, T.; Guimarães-Camboa, N.; Banerjee, I.; Zambon, A. C.; Kisseleva, T.;
Velayoudon, A.; Stallcup, W. B.; Gu, Y.; Dalton, N. D.; Cedenilla, M.; et al. Resident fibroblast
lineages mediate pressure overload-induced cardiac fibrosis. J Clin Invest 2014, 124 (7), 2921-
2934. DOI: 10.1172/JCI74783.
(287) Ali, S. R.; Ranjbarvaziri, S.; Talkhabi, M.; Zhao, P.; Subat, A.; Hojjat, A.; Kamran, P.; Müller,
A. M.; Volz, K. S.; Tang, Z.; et al. Developmental heterogeneity of cardiac fibroblasts does not
predict pathological proliferation and activation. Circ Res 2014, 115 (7), 625-635. DOI:
10.1161/CIRCRESAHA.115.303794.
(288) Zeisberg, E. M.; Tarnavski, O.; Zeisberg, M.; Dorfman, A. L.; McMullen, J. R.; Gustafsson,
E.; Chandraker, A.; Yuan, X.; Pu, W. T.; Roberts, A. B.; et al. Endothelial-to-mesenchymal
transition contributes to cardiac fibrosis. Nat Med 2007, 13 (8), 952-961. DOI: 10.1038/nm1613.
(289) Tillmanns, J.; Hoffmann, D.; Habbaba, Y.; Schmitto, J. D.; Sedding, D.; Fraccarollo, D.;
Galuppo, P.; Bauersachs, J. Fibroblast activation protein alpha expression identifies activated
fibroblasts after myocardial infarction. J Mol Cell Cardiol 2015, 87, 194-203. DOI:
10.1016/j.yjmcc.2015.08.016.
(290) Arnold, J. N.; Magiera, L.; Kraman, M.; Fearon, D. T. Tumoral immune suppression by
macrophages expressing fibroblast activation protein-α and heme oxygenase-1. Cancer Immunol
Res 2014, 2 (2), 121-126. DOI: 10.1158/2326-6066.CIR-13-0150.
(291) Mori, Y.; Kono, K.; Matsumoto, Y.; Fujii, H.; Yamane, T.; Mitsumata, M.; Chen, W. T. The
expression of a type II transmembrane serine protease (Seprase) in human gastric carcinoma.
Oncology 2004, 67 (5-6), 411-419. DOI: 10.1159/000082926.

131
(292) Kelly, T.; Kechelava, S.; Rozypal, T. L.; West, K. W.; Korourian, S. Seprase, a membranebound protease, is overexpressed by invasive ductal carcinoma cells of human breast cancers.
Mod Pathol 1998, 11 (9), 855-863.
(293) Aghajanian, H.; Kimura, T.; Rurik, J. G.; Hancock, A. S.; Leibowitz, M. S.; Li, L.; Scholler,
J.; Monslow, J.; Lo, A.; Han, W.; et al. Targeting cardiac fibrosis with engineered T cells. Nature
2019, 573 (7774), 430-433. DOI: 10.1038/s41586-019-1546-z.
(294) Frey, N.; Porter, D. Cytokine Release Syndrome with Chimeric Antigen Receptor T Cell
Therapy. Biol Blood Marrow Transplant 2019, 25 (4), e123-e127. DOI:
10.1016/j.bbmt.2018.12.756.
(295) Pan, Q.; Xu, J.; Wen, C. J.; Xiong, Y. Y.; Gong, Z. T.; Yang, Y. J. Nanoparticles: Promising
Tools for the Treatment and Prevention of Myocardial Infarction. Int J Nanomedicine 2021, 16,
6719-6747. DOI: 10.2147/IJN.S328723.
(296) Somasuntharam, I.; Yehl, K.; Carroll, S. L.; Maxwell, J. T.; Martinez, M. D.; Che, P. L.;
Brown, M. E.; Salaita, K.; Davis, M. E. Knockdown of TNF-α by DNAzyme gold nanoparticles as
an anti-inflammatory therapy for myocardial infarction. Biomaterials 2016, 83, 12-22. DOI:
10.1016/j.biomaterials.2015.12.022.
(297) Maranhão, R. C.; Guido, M. C.; de Lima, A. D.; Tavares, E. R.; Marques, A. F.; Tavares de
Melo, M. D.; Nicolau, J. C.; Salemi, V. M.; Kalil-Filho, R. Methotrexate carried in lipid core
nanoparticles reduces myocardial infarction size and improves cardiac function in rats. Int J
Nanomedicine 2017, 12, 3767-3784. DOI: 10.2147/IJN.S129324.
(298) Chung, E. J. Targeting and therapeutic peptides in nanomedicine for atherosclerosis. Exp
Biol Med (Maywood) 2016, 241 (9), 891-898. DOI: 10.1177/1535370216640940.
(299) Ma, S.; Tian, X. Y.; Zhang, Y.; Mu, C.; Shen, H.; Bismuth, J.; Pownall, H. J.; Huang, Y.;
Wong, W. T. E-selectin-targeting delivery of microRNAs by microparticles ameliorates endothelial
inflammation and atherosclerosis. Sci Rep 2016, 6, 22910. DOI: 10.1038/srep22910.
(300) Chin, D. D.; Poon, C.; Trac, N.; Wang, J.; Cook, J.; Joo, J.; Jiang, Z.; Maria, N. S. S.; Jacobs,
R. E.; Chung, E. J. Collagenase-Cleavable Peptide Amphiphile Micelles as a Novel Theranostic
Strategy in Atherosclerosis. Adv Ther (Weinh) 2020, 3 (3). DOI: 10.1002/adtp.201900196.
(301) Li, Y.; Xuan, J.; Hu, R.; Zhang, P.; Lou, X.; Yang, Y. Microfluidic triple-gradient generator
for efficient screening of chemical space. Talanta 2019, 204, 569-575. DOI:
10.1016/j.talanta.2019.06.018.

132
(302) Frick, C.; Dettinger, P.; Renkawitz, J.; Jauch, A.; Berger, C. T.; Recher, M.; Schroeder, T.;
Mehling, M. Nano-scale microfluidics to study 3D chemotaxis at the single cell level. PLoS One
2018, 13 (6), e0198330. DOI: 10.1371/journal.pone.0198330.
(303) Gandavarapu, N. R.; Azagarsamy, M. A.; Anseth, K. S. Photo-click living strategy for
controlled, reversible exchange of biochemical ligands. Adv Mater 2014, 26 (16), 2521-2526. DOI:
10.1002/adma.201304847.
(304) Hausenloy, D. J.; Bøtker, H. E.; Ferdinandy, P.; Heusch, G.; Ng, G. A.; Redington, A.;
Garcia-Dorado, D. Cardiac innervation in acute myocardial ischaemia/reperfusion injury and
cardioprotection. Cardiovasc Res 2019, 115 (7), 1167-1177. DOI: 10.1093/cvr/cvz053.
(305) Gentek, R.; Hoeffel, G. The Innate Immune Response in Myocardial Infarction, Repair, and
Regeneration. Adv Exp Med Biol 2017, 1003, 251-272. DOI: 10.1007/978-3-319-57613-8_12.
(306) Stevens, K. R.; Ungrin, M. D.; Schwartz, R. E.; Ng, S.; Carvalho, B.; Christine, K. S.;
Chaturvedi, R. R.; Li, C. Y.; Zandstra, P. W.; Chen, C. S.; et al. InVERT molding for scalable
control of tissue microarchitecture. Nat Commun 2013, 4, 1847. DOI: 10.1038/ncomms2853.
(307) Hoang, P.; Wang, J.; Conklin, B. R.; Healy, K. E.; Ma, Z. Generation of spatial-patterned
early-developing cardiac organoids using human pluripotent stem cells. Nat Protoc 2018, 13 (4),
723-737. DOI: 10.1038/nprot.2018.006.
(308) Nikolaev, M.; Mitrofanova, O.; Broguiere, N.; Geraldo, S.; Dutta, D.; Tabata, Y.; Elci, B.;
Brandenberg, N.; Kolotuev, I.; Gjorevski, N.; et al. Homeostatic mini-intestines through scaffoldguided organoid morphogenesis. Nature 2020, 585 (7826), 574-578. DOI: 10.1038/s41586-020-
2724-8.
(309) Mirdamadi, E.; Tashman, J. W.; Shiwarski, D. J.; Palchesko, R. N.; Feinberg, A. W. FRESH
3D Bioprinting a Full-Size Model of the Human Heart. ACS Biomater Sci Eng 2020, 6 (11), 6453-
6459. DOI: 10.1021/acsbiomaterials.0c01133.
(310) Skylar-Scott, M. A.; Uzel, S. G. M.; Nam, L. L.; Ahrens, J. H.; Truby, R. L.; Damaraju, S.;
Lewis, J. A. Biomanufacturing of organ-specific tissues with high cellular density and embedded
vascular channels. Sci Adv 2019, 5 (9), eaaw2459. DOI: 10.1126/sciadv.aaw2459.
(311) Bittner, S. M.; Smith, B. T.; Diaz-Gomez, L.; Hudgins, C. D.; Melchiorri, A. J.; Scott, D. W.;
Fisher, J. P.; Mikos, A. G. Fabrication and mechanical characterization of 3D printed vertical
uniform and gradient scaffolds for bone and osteochondral tissue engineering. Acta Biomater
2019, 90, 37-48. DOI: 10.1016/j.actbio.2019.03.041.

133
(312) Liu, W.; Zhang, Y. S.; Heinrich, M. A.; De Ferrari, F.; Jang, H. L.; Bakht, S. M.; Alvarez, M.
M.; Yang, J.; Li, Y. C.; Trujillo-de Santiago, G.; et al. Rapid Continuous Multimaterial Extrusion
Bioprinting. Adv Mater 2017, 29 (3). DOI: 10.1002/adma.201604630.
(313) Yang, X.; Pabon, L.; Murry, C. E. Engineering adolescence: maturation of human pluripotent
stem cell-derived cardiomyocytes. Circ Res 2014, 114 (3), 511-523. DOI:
10.1161/CIRCRESAHA.114.300558.
(314) Dai, X.; Wiernek, S.; Evans, J. P.; Runge, M. S. Genetics of coronary artery disease and
myocardial infarction. World J Cardiol 2016, 8 (1), 1-23. DOI: 10.4330/wjc.v8.i1.1.
(315) Zhao, Y.; Wang, E. Y.; Davenport, L. H.; Liao, Y.; Yeager, K.; Vunjak-Novakovic, G.;
Radisic, M.; Zhang, B. A Multimaterial Microphysiological Platform Enabled by Rapid Casting of
Elastic Microwires. Adv Healthc Mater 2019, 8 (5), e1801187. DOI: 10.1002/adhm.201801187.
(316) Yip, J. K.; Sarkar, D.; Petersen, A. P.; Gipson, J. N.; Tao, J.; Kale, S.; Rexius-Hall, M. L.;
Cho, N.; Khalil, N. N.; Kapadia, R.; et al. Contact photolithography-free integration of patterned
and semi-transparent indium tin oxide stimulation electrodes into polydimethylsiloxane-based
heart-on-a-chip devices for streamlining physiological recordings. Lab Chip 2021. DOI:
10.1039/d0lc00948b.
(317) Dipalo, M.; Rastogi, S. K.; Matino, L.; Garg, R.; Bliley, J.; Iachetta, G.; Melle, G.; Shrestha,
R.; Shen, S.; Santoro, F.; et al. Intracellular action potential recordings from cardiomyocytes by
ultrafast pulsed laser irradiation of fuzzy graphene microelectrodes. Sci Adv 2021, 7 (15). DOI:
10.1126/sciadv.abd5175.
(318) Marasco, C. C.; Enders, J. R.; Seale, K. T.; McLean, J. A.; Wikswo, J. P. Real-time cellular
exometabolome analysis with a microfluidic-mass spectrometry platform. PLoS One 2015, 10 (2),
e0117685. DOI: 10.1371/journal.pone.0117685.
(319) McKenzie, J. R.; Cognata, A. C.; Davis, A. N.; Wikswo, J. P.; Cliffel, D. E. Real-Time
Monitoring of Cellular Bioenergetics with a Multianalyte Screen-Printed Electrode. Anal Chem
2015, 87 (15), 7857-7864. DOI: 10.1021/acs.analchem.5b01533.
(320) Zhang, Y. S.; Aleman, J.; Shin, S. R.; Kilic, T.; Kim, D.; Mousavi Shaegh, S. A.; Massa, S.;
Riahi, R.; Chae, S.; Hu, N.; et al. Multisensor-integrated organs-on-chips platform for automated
and continual in situ monitoring of organoid behaviors. Proc Natl Acad Sci U S A 2017, 114 (12),
E2293-E2302. DOI: 10.1073/pnas.1612906114.
(321) Kim, I.; Choi, S.; Yoo, S.; Lee, M.; Kim, I. S. Cancer-Associated Fibroblasts in the Hypoxic
Tumor Microenvironment. Cancers (Basel) 2022, 14 (14). DOI: 10.3390/cancers14143321.

134

Abstract (if available)

Abstract A myocardial infarction, or heart attack, disrupts the flow of vital, oxygenated blood to downstream tissue, leading to cardiac myocyte necrosis within hours. Because the mammalian heart has a limited capacity to regenerate, the subsequent wound healing process relies on the formation of scar tissue. Cardiac fibroblasts play a critical role in this process by activating into a myofibroblast phenotype, which deposits matrix that ultimately forms a scar. Although this is necessary to maintain tissue integrity, excessive fibrosis often occurs and is a common feature of heart failure. Whether crosstalk between hypoxic, injured tissue and neighboring, healthy myocardium may regulate mechanisms of fibrosis has been relatively unexplored. This is due largely to shortcomings of existing research tools, which offer limited spatiotemporal control over relevant biophysical features of the cell microenvironment. Organs on Chips have recently emerged to recapitulate structural and functional features of human tissue, such as oxygen level. Our objective was to develop a Post-Infarct Myocardium on a Chip for the co-culture of hypoxic and normoxic cardiac cells. We first characterized human cardiac fibroblast phenotypes in acute hypoxia. We then engineered and implemented a Post-Infarct Myocardium on a Chip for oxygen control to identify crosstalk between hypoxic and normoxic fibroblasts. Lastly, we adapted it for the heterotypic co-culture of hypoxic fibroblasts and normoxic cardiac myocytes. We find that oxygen-dependent crosstalk exacerbates pro-inflammatory signaling. This work contributes to Organ on Chip models of post-infarct myocardium and has applications in disease modeling and drug screening.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Extracellular matrix regulation of mitochondrial function in engineered cardiac myocytes

PDF

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

PDF

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

PDF

Investigating the role of cardiac fibroblasts in modulating angiogenesis during heart injury

Asset Metadata

Creator Khalil, Natalie (author)

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

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Biomedical Engineering

Degree Conferral Date 2024-05

Publication Date 12/12/2024

Defense Date 05/02/2024

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

Tag cardiac fibrosis,crosstalk,hypoxia,myocardial infarction,OAI-PMH Harvest,organ-on-a-chip

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor McCain, Megan (committee chair), Chung, Eun Ji (committee member), Lien, Ching-Ling (committee member), Mumenthaler, Shannon (committee member), Shen, Keyue (committee member)

Creator Email khalil.natalie@gmail.com,nnkhalil@usc.edu

Unique identifier UC1139963Q0

Identifier etd-KhalilNata-13080.pdf (filename)

Legacy Identifier etd-KhalilNata-13080

Document Type Dissertation

Format theses (aat)

Rights Khalil, Natalie

Internet Media Type application/pdf

Type texts

Source 20240614-usctheses-batch-1168 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright.

Repository Name University of Southern California Digital Library

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

Repository Email cisadmin@lib.usc.edu