University of Southern California Dissertations and Theses

Derivation of cardiomyocyte-propelled motile cellular constructs via de novo developmental trajectory

(USC Thesis Other)

Derivation of cardiomyocyte-propelled motile cellular constructs via de novo developmental trajectory

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content DERIVATION OF CARDIOMYOCYTE-PROPELLED CELLULAR CONSTRUCTS VIA DE
NOVO DEVELOPMENTAL TRAJECTORY
by
Gaveen Godage
A Thesis Presented to the
FACULTY OF THE USC KECK SCHOOL OF MEDICINE
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(STEM CELL BIOLOGY AND REGENERATIVE MEDICINE)
December 2024

Acknowledgments
Foremost, I would like to express my deepest gratitude to my thesis advisor and
committee chair, Dr. Leonardo Morsut. With only curiosity to offer, Dr. Morsut graciously took
me into the lab as a young undergraduate. Your guidance and encouragement have been
instrumental to this project and my development as a young scientist. I am also grateful for your
humility and thoughtful leadership, and always taking the time to ask about life outside of
science.
Next, I would like to thank fellow committee members and mentors Dr. Francesca
Mariani and Dr. Qi-Long Ying. The opportunity to pursue this thesis would not have been
possible without the support and flexibility of Dr. Mariani. Thank you for your endless
commitment to student success. The stories and accomplishments of myself and many other
students in the department can not be told without you as a central figure. Thank you Dr. Ying
for your insightful advice, critical questioning, and dedication to experimental excellence. All
members of the committee have provided meaningful feedback and tangible implementations.
I would also like to thank all of the members of the Morsut Lab, it was a privilege to
work alongside such an integrated and brilliant team. A great thanks to Dr. Christine Ho, Dr.
Benjamin Swedlund, Dr. Fokion Glyforkydis, and Kyle Poon who all aided in the progression of
this project. A special thank you to Dr. Ho who took the time to train me as she was not only
balancing her science, but also her move to the U.S.
Lastly, I would like to thank my family for their unconditional support and
encouragement throughout this project. They are the cornerstone of my efforts and have instilled
a belief to dream ambitiously and work with tenacity.
Clearly, this accomplishment would not have been possible without many and is a
reflection of the gracious support I have been fortunate to receive.
ii

Table of Contents
Acknowledgements........................................................................................................................ ii
List of Figures.................................................................................................................................v
Abstract..........................................................................................................................................vi
Chapter 1: Introduction................................................................................................................ 1
1.1 Embryonic Development..................................................................................................... 1
1.2 Biohybrid Robotics: Engineering Tissue for Function........................................................ 1
1.3 Stem Cells as a Tissue Model in vitro..................................................................................2
1.4 Cardiac Tissue Morphogenesis............................................................................................ 3
1.5 Synthetic Biology: Applications in Tissue Engineering......................................................6
1.6 Objectives of this Thesis......................................................................................................8
Chapter 2: Materials and Methods............................................................................................ 10
2.1 Cell Lines and Maintenance...............................................................................................10
2.2 Gastruloid, Cardiogenic Gastruloid and ABV_CDM Culture...........................................10
2.3 Live Imaging and Image Analysis..................................................................................... 12
2.4 Calcium Imaging and Dynamics........................................................................................13
2.5 Motility Assay....................................................................................................................13
2.6 Motility and Morphological Feature Analysis...................................................................14
2.6 Cadherin Cell Line Generation.......................................................................................... 15
2.7 Doxycycline Time Course Experiment..............................................................................18
2.8 Flow Cytometry Analysis.................................................................................................. 18
2.9 Immunofluorescence..........................................................................................................19
2.10 Statistical Analysis...........................................................................................................20
Chapter 3: Results........................................................................................................................21
3.1.1 Capacity for current organoid protocols to generate contractile aggregates...................21
3.1.2 Effect of ABV_CDM on aggregate area, eccentricity, and length..................................21
iii

3.1.3 Effect of ABV_CDM on efficiency of deriving contractile aggregates on Day 6 and
Day 7........................................................................................................................................23
3.1.4 Effect of ABV_CDM on percentage area of Contraction and calcium dynamics..........23
3.2.1 Effect of ABV_CDM on Flk1 and PDGRFα cardiac mesoderm markers......................24
3.2.2 Effect of ABV_CDM on Nkx2.5 expression..................................................................24
3.2.3 Effect of ABV_CDM on Brachyury and Sox-17 expression..........................................25
3.2.4 Effect of ABV_CDM on germ layer distribution at Day 5.............................................25
3.2.5 Effect of ABV_CDM on emergence of motility at Day 10............................................ 26
3.2.6 Associating morphological and functional features with motility..................................27
3.3.1 Design of cadherin overexpressing cell lines..................................................................28
3.3.2 FACS and immunofluorescence analysis for cadherin overexpression after 24 hour
doxycycline induction..............................................................................................................28
3.3.3 Effect of prolonged doxycycline induction response on mCherry signal.......................29
Chapter 4: Discussion..................................................................................................................30
4.1 Impact of temporal delivery of soluble factors on tissue developmental trajectories........30
4.2 Morphological properties associated with motility............................................................32
4.3 Challenges of overexpressing cadherins in ECSs..............................................................33
Chapter 5: Tables and Figures....................................................................................................36
References.....................................................................................................................................59
iv

List of Figures
Figure 1. Current models of myocyte powered biohybrid robots................................................. 36
Figure 2. Schematic of Activin/Nodal and BMP Signaling..........................................................37
Figure 3. Schematic of ZX1 Inducible Cassette Exchange and Tetracycline-On System. .......... 38
Figure 4. Currently available protocols for 3D Gastruloids and Cardiogenic Gastruloids...........39
Figure 5. ABV_CDM differentiation protocol for motile aggregates...........................................40
Figure 6. Morphological features of aggregates generated across protocols................................41
Figure 7. Percent of contractile aggregates on day 6 and day 7 of differentiation....................... 42
Figure 8. Calcium dynamics and contractile area percentages of aggregates...............................43
Figure 9. FACS analysis of Flk1+PDGRFα+ double positive populations across protocols......... 45
Figure 10. FACS analysis of Nkx2.5+ positive population across protocols................................ 46
Figure 11. FACS analysis of Brachyury+ and Sox-17+ positive populations across protocols..... 47
Figure 12. Paired expression of cell markers during ABV_CDM differentiation protocol..........48
Figure 13. Immunofluorescence staining for germ layers on Day 5.............................................49
Figure 14. Batch motility assay and analysis on Day 10.............................................................. 50
Figure 15. Morphological, contractile, and motility feature analysis at individual aggregate level.
……………………………………………………………………………………………………51
Figure 16. Plasmid design for generated cell lines with ZX1 inducible cassette exchange system.
……………………………………………………………………………………………………52
Figure 17. FACS analysis for cadherin expression after 24 hour doxycycline induction.............53
Figure 18. Immunofluorescence staining for N-Cadherin overexpression after 24 hour
doxycycline induction....................................................................................................................54
Figure 19. Immunofluorescence staining for P-Cadherin overexpression after 24 hour
doxycycline induction....................................................................................................................55
Figure 20. Immunofluorescence staining for E-Cadherin overexpression after 24 hour
doxycycline induction....................................................................................................................56
Figure 21. FACS analysis for mCherry signal with increasing duration of doxycycline induction.
……………………………………………………………………………………………………57
v

Abstract
Tissue engineering and biorobotics propel our understanding of cellular coordination,
multicellular assembly, and the interplay between form and function. Efforts have probed the
capacity to use cellular components as actuators, sensors, and structural frames, integrating them
with non-living material to create bio-hybrid machines. Of particular interest is the development
of microswimmers, constructs that can swim at small scales with applications in drug delivery,
bio-electronics, and ecological exploration. Recently, efforts have explored the ability to
engineer cellular-only constructs of microswimmers. The work aims to produce novel anatomic
configurations and emergent functions from the intrinsic machinery of cells. Thus, pushing our
knowledge of cell plasticity and collective behavior. Several successful attempts have been made
in this pursuit. Most recently, Gumuskaya et al., 2023 published a protocol for Anthrobots. These
are self-organizing and cilia powered biobots of stem cells derived from adult human lung tissue.
This thesis puts forth a new player in the realm of biobots and biological engineering: a
self-motile cellular construct via cardiomyocyte actuation. Unlike previous constructs, this
protocol entails the use of mouse embryonic stem cells that are coaxed towards a novel body
plan and myocyte-powered motility. The protocol is 10 days long, with emergence of
contractility and motility by days 6 and 7 respectively, and consists of soluble factor delivery
during specific temporal windows. The final constructs are a mean 1019±101µm in length. In a
motility assay on day 10 they display a mean displacement of 788.67±289.56µm. Further, I
discuss developmental marker dynamics and morphological features that comprise the motile
aggregates. Finally, I present an attempt to synthetically engineer aggregate body plans through
cadherin mediated phase separation, with discussion on its pitfalls and consideration for future
implementations.
vi

Chapter 1: Introduction
1.1 Embryonic Development
From embryonic to adult form nature commands a simply incomprehensible coordination
of biology to form tissues, a process known as tissue morphogenesis (Wu et al., 2023; Morsut
2017). Starting as undifferentiated and amorphous, single cells of the metazoa are able to
self-organize into multicellular tissues of diverse structure and function. Cells make their
developmental decisions based on the mechanical and chemical stimuli from neighboring cells
and the extracellular matrix (Heisenberg and Bellaïche 2013; Swartz 2003). A few of these
responses include: morphological adaptations, division, differentiation, secretion, migration, or
even death (Zallen and Goldstein 2017). In turn, these mechanical and physiological outputs alter
their environment. The emergent function and final architecture of a tissue are a result of this
collaboration between individual cells. How are cells able to form tissues? And, can we harness
this ability for therapeutic purposes? These are fundamental questions of developmental biology
and regenerative medicine. The work of this thesis aims to contribute to both.
1.2 Biohybrid Robotics: Engineering Tissue for Function
Awed by the elegance of natural tissue organization, scientists have attempted to build
tissues in the lab. The field of biological robotics entails interfacing engineering technologies
with biology to create hybrid systems (Blackiston et al., 2023). Technologies such as 3D
micro-patterning or material scaffolding are used to control collective geometry and individual
cell placement in order to build living machines that mimic natural biological systems or have
novel functions (Pagan-Diaz et al., 2018). One particular interest in the field is the construction
of microswimmers, constructs that can swim at small scales, which have applications in drug
1

delivery, surgery, bio-electronics, and therapeutics (Bunea and Taboryski 2020; Mo et al., 2023).
Several iterations of microswimmers have been published with various propulsion mechanisms.
One seminal model in the field was published by Nawroth et al. 2012. They constructed a freely
moving microswimmer dubbed “medusoids” that consisted of rat cardiomyocytes, the contractile
cells of the heart, placed along a silicone polymer scaffold. Inspired by the rhythmic swimming
dynamics of jellyfish, the scaffold was designed to mimic their natural body plan. Pacemaker
cells were strategically placed along the body plan so that the waves generated by
cardiomyocytes induced body deformations conducive to swimming (Fig. 1a). This work was a
major achievement in the field of soft-robotics, demonstrating the ability to reverse engineer
function and form of muscular organs and biological swimmers. Since then there have been
many others including: pumps, grasping machines, slithering worms, and biohybrid fish (Fig.
1b,c) (Yuan et al., 2023; Young Lee et al., 2022). This work also demonstrated the potential for
cell based actuation via cardiomyocytes. Although remarkable, these examples always entail
protocol of classic fabrication where different materials are put together by an external user; to
date, no biorobot that is able to construct itself, made only of cells, has generated a
microswimmer via cardiomyocyte actuation. I aim to do so by harnessing the intrinsic
capabilities of stem cells.
1.3 Stem Cells as a Tissue Model in vitro
Stem cells are cells that have the innate ability to self-renew or differentiate into one or
many cell types. Pluripotent stem cells (PSCs) are a subset of stem cells that have the ability to
give rise to all of the cell types of an organism. Embryonic stem cells (ESCs) are a type of PSCs
that are derived from the inner cell mass (ICM) of the blastocyst, an early stage embryo
2

(Thomson et al., 1998). ESCs were first derived in 1980 and are contemporarily cultured with
leukemia inhibitory factor with MEK and GSK3 inhibition (LIF + 2i) to maintain their
pluripotency in vitro (Efroni et al., 2008; Ying et al., 2008). During development, ESCs can
undergo lineage specification whereby they are directed toward a specific cell fate and lose their
pluripotency. This process can be recapitulated in vitro by supplying cells with the signaling
molecules and environmental perturbations that mimic the regulatory pathways corresponding to
that lineage establishment during embryogenesis. Therefore, ESCs serve as a powerful tool to
model tissue formation and delineate developmental trajectories. Whether stem cells can be
coaxed to construct non-native developmental trajectories has started to be explored. Recently,
Gumuskaya et al. 2024 developed a protocol to drive the self-organization of adult stem cells
into motile biobots deemed anthrobots. Anthrobots are spheroid-shaped aggregates with
cilia-driven propulsion. Whether we can generate cardiogenic-mediated motility via ESCs is not
known.
1.4 Cardiac Tissue Morphogenesis
Cardiac tissue differentiation from ESCs has been of particular interest for disease
modeling, drug toxicity testing, and therapeutic interventions (Kattman et al., 2011). Therefore,
substantial efforts have been made to model cardiac tissue formation in vitro. Below are relevant
signaling pathways and their implementation in cardiac differentiation in vivo and in vitro.
1.4.1 Roles of Activin/Nodal and BMP Signaling Pathways in Cardiogenesis
3

Two key regulation pathways involved during embryonic cardiogenesis are the
Activin/Nodal and Bone Morphogenetic Protein-4 pathways, both of which are a part of the
TGF-β family of morphogens (Pauklin and Vallier 2015; Wang et al., 2014).
Activin and Nodal both act on transmembrane Activin Type I and II serine/threonine
kinase receptors. Upon binding, these receptors phosphorylate Smad2 and Smad3, which then
complex with Smad4 and translocate to the nucleus to regulate transcription through epigenetic
modifiers at target loci (Fig. 2) (Pauklin and Vallier 2015). During embryonic development, the
described Activin/Nodal pathway plays an essential role in the induction of the primitive streak,
which establishes the anterior-posterior axis and gives rise to definitive endoderm and
mesoderm. This pathway further directs some mesoderm into cardiac mesoderm and gives rise to
the cardiac lineage (Papanayotou and Collignon 2014).
Bone morphogenetic proteins (BMPs) are named based on the initial discovery of their
role in bone and cartilage formation. However, they play an important role in several tissues,
including the heart. BMPs bind to a wide array of receptors in the TGF-β pathway and form
tetrameric complexes with type I and type II transmembrane serine/threonine receptors. Upon
binding, these receptors phosphorylate the BMP associated Smads, which are Smad1, Smad5,
and Smad8 (Wang et al., 2014). These Smads then complex with Smad4 and translocate to the
nucleus to regulate transcription (Fig. 2). During embryonic development, BMPs play a
prominent role in ventralization of activin-induced mesoderm and cardiac development.
Specifically, BMP2 and BMP4 are required for cardiac progenitor specification and later
induction of differentiation. Knockout studies of BMPs and/or their receptors have led to the
absence of mesoderm and severe cardiac malformations (Wang et al., 2011).
4

1.4.2 Stage Specific Cardiac Differentiation of ESCs in 2D
In order to efficiently differentiate ESCs to cardiomyocytes, the contractile cells of the
heart, in vitro optimal levels of Activin/Nodal and BMP signaling are needed. Several studies
have contributed to our understanding of the developmental trajectory of cardiac lineage
specification. In a landmark study, Kattman et al. 2011, highlighted the dosage and temporal
windows needed for proper induction. To do so, they first identified markers for the presence of
cardiac mesoderm and subsequent differentiation. Flk-1+ PDGFR-α+ double positive cells were
shown to mark cardiac mesoderm with cardiomyocyte potential and Nkx2.5 was shown to mark
cardiac precursors. Using these markers, they were able to quantitatively measure and track
mesoderm induction, allowing for proper temporal delivery of Activin A and BMP4 to induce
cardiac differentiation. Along with temporal considerations, varying titrations of factors were
needed across cell lines based on differing endogenous signaling. Overall, the study highlighted
the ability to design and standardize stage-specific differentiation of ESCs into cardiomyocytes
in vitro.
1.4.3 Embryogenesis and Cardiogenesis in 3D
The protocol that we are aiming to develop is one where cells are self-organizing by
themselves, with as little intervention from the outside as possible. Recently, stem cell derived
organoids have been developed: organoids are 3D self-organizing miniature structures that
mimic the cellular composition, architecture, and even function of tissues found in organs (Yang
et al., 2023). Embryonic organoids have been shown to recapitulate embryonic developmental
events such as, pre-implantation blastocyst, post-implantation development, gastrulation, and
most recently even early organogenesis (van den Brink and Oudenaarden 2021). A protocol for
5

the derivation of gastrulation, the emergence of the three germ layers, was developed by van den
Brink et al., 2014. They show that the activation of Wnt signaling, via small molecule Chiron, in
ESC embryoid bodies is able to induce derivatives of all three germ layers patterned along the
anterior-posterior axis. These structures were deemed gastruloids. Building on this development,
a protocol supporting embryonic cardiac development was published by Rossi et al., 2021. Here
they followed the gastruloid protocol, but added a cocktail of cardiogenic factors during a
specific temporal window to induce cardiogenesis. These structures were then found to
differentiate into functional cardiomyocytes at the anterior pole. These were named cardiogenic
gastruloids. These efforts highlight the remarkable ability to direct complex self-organization of
ESCs in vitro and recapitulate key in vivo developmental events. The organoid field continues to
lead tremendous efforts in furthering these capabilities. However, it is still yet to be explored
whether or not we can exert control over these self-organizing capabilities to generate tissues
with spatially defined patterns or novel emergent functions, such as motility.
1.5 Synthetic Biology: Applications in Tissue Engineering
Broadly, synthetic biology is the application of systems engineering to biological
constructs. One aspect of this field is genomic modification. Gene editing is a powerful tool to
investigate gene function, an especially pertinent tool for studying cell fate specification and
tissue morphogenesis (Gaj et al., 2016). Current methods of gene editing include Zinc Finger
Nucleases (ZFNs), Transcription Activator-like Effector Nucleases (TALENs), and Clustered
Regularly Interspaced Short Palindromic Repeats-Cas9 (CRISPR-Cas9) system. These tools
enable the modification of DNA in a targeted manner. ZFNs and TALENs require engineering
protein domains that bind to the DNA sequence of interest (Boch et al., 2009; Kim et al., 1996).
6

CRISPR-Cas9 entails engineering a guide RNA that binds to the complementary DNA sequence
of interest (Wiedenheft et al., 2012). Once bound to the target site double stranded breaks (DSBs)
are induced using a nuclease, an enzyme that cuts bonds between nucleotides. ZNFs and
TALENs induce DSBs using the FokI nuclease, derived from the bacterium Flavobacterium
okeanokoites. CRISPR-Cas9 induce DSBs using the Cas9 nuclease, derived from Streptococcus
pyogenes. Following the DNA break, endogenous mechanisms repair the DNA.
Non-Homologous End Joining (NHEJ) repairs DSBs by directly ligating the broken strands
together without a template strand. Homology-Directed Repair (HDR) repairs DSBs using a
homologous DNA template.
These systems can be used to induce permanent deletions or insertions within the
genome, providing a method for gene knockout or mutation studies. Furthermore, these methods
can be used to overexpress genes of interest. In order to improve the rates of homologous
recombination and decrease off target mutations due to nucleases, Lacovino et al., 2011
developed the ZX1 Inducible Cassette Exchange System.
1.5.1 ZX1 Inducible Cassette Exchange System
The ZX1 Inducible Cassette Exchange System (ICE) is a recombination system that
allows for the insertion of a gene cassette into a predefined locus under conditional control in
ESCs. The modified ESC cell line has a Cre-recombinase flanked by LoxP sites under the
Tetracycline-Responsive Element (TRE) promoter. Cre-recombinase is an enzyme, derived from
bacteriophage P1, that recognizes LoxP sites within a genome and deletes the DNA sequence
between them. This cassette is located on the X chromosome in the genomic region upstream of
the hypoxanthine phosphoribosyltransferase (HPRT) locus, which has been shown to express
7

transgenes reliably (Bronson et al., 1996; Touw et al., 2007). Plasmids are designed with the
gene of interest flanked with LoxP sites. Upon transfection, a single copy of the transgene is
recombined and integrated into the ICE locus. The engineered cell line, plasmid, and ICE
product are shown in Figure 3a.
The conditional expression of the gene is through the Tetracycline On system. This
system consists of the Reverse Tetracycline-Controlled Transactivator (rtTA), a genetically
engineered fusion protein that binds to TRE. In the absence of tetracycline or doxycycline rtTA
remains unbound to TRE and transcription is repressed. Upon the addition to the culture media,
tetracycline binds to rtTA and activates it to bind to TRE and initiate transcription (Fig. 3b) (Das
et al., 2016). The system can be switched off by replenishing fresh medium. Thereby, allowing
for conditional control over gene expression. The ZX1 ICE system has an inserted rtTA on the
ROSA26 safe harbor locus on chromosome 6. I utilized this system to conditionally overexpress
genes of interest in ESCs.
1.6 Objectives of this Thesis
The work of this thesis aims to further explore the possibility of coaxing cells to
self-construct into tissues with user-defined structures and functions. Inspired by both the
classical biofabrication of microswimmers and novel 3D organoid cultures, I aim to integrate
foundations of each to develop a cell-only microswimmer construct that is motile via
cardiomyocyte actuation.
The first aim is the optimization of a protocol for the generation of mESC-derived
elongated and contractile cellular constructs via a soluble factor approach. This approach will
consist of the delivery of cardiogenic inducing factors, such as Activin A, BMP4 and VEGF, to
8

increase the differentiation efficiency and contractility in 3D gastruloid models. The strategy
employed will be testing factor delivery during non-canonical temporal windows.
The second aim of this work is the characterization of the developmental trajectory and
emergent function of these constructs. To do so, I conduct flow cytometry and
immunofluorescence analysis of cardiac lineage markers. Further, I assess the capacity for
motility and assess correlations between morphological and functional features.
The third aim of this work is to utilize the ZX1 ICE system to engineer greater control of
tissue patterning and body plan. To do so, I explain a plan to engineer ESCs to conditionally
overexpress adhesion proteins, such as cadherins, to drive phase separation in 3D multicellular
constructs.
9

Chapter 2: Materials and Methods
2.1 Cell Lines and Maintenance
All mESC lines were maintained at 37℃, 5% CO2
in DMEM (GIBCO-11710035)
supplemented with 10% Embryonic Stem Cell qualified FBS, 2mM L-Glutamine
(GIBCO-A2916801), 1mM Sodium Pyruvate (GIBCO-11360070), 0.1mM β-mercaptoethanol
(Sigma M3148), 100U/mL Penicillin-Streptomycin (GIBCO-15140122), 1:100 MEM
Non-Essential Amino Acids (GIBCO-11140-050), 1x Leukemia inhibitory factor (LIF, Santa
Cruz sc-4989), 3µM GSK3 inhibitor (CHIR99021 Stem Cell Technologies, and 1µM MEK
inhibitor (PD0325901 Selleckchem S1036). For passaging, cells were rinsed with PBS
(GIBCO-10010023) and detached via 5 minute incubation at 37℃, 5% CO2 with StemPro
accutase (GIBCO-A11105), blocked with DMEM and 10% FBS, centrifuged for 5 minutes at
300 rpm, and resuspended in maintenance media. Cells were passed around 70-80% confluence,
every other day, and replated in a 1% Gelatin coated 6-well plate (Falcon-353046). The ES 14
cell lines used were: Tbra-eGFP (Fehling et al., 2003), Nkx2.5-eGFP (Hsia et al., 2008),
Sox-17-RFP,Bra-GFP (Pour et al., 2022). Aim 3 used cell lines: P2Lox-mCherry-Ncad,
P2lox-mCherry-Ecad, P2Lox-mCherry-Pcad, and P2Lox-CAG-GFP, from the parent ZX1 cell
line (Lacaovino et al., 2011) and have a detailed generation below.
2.2 Gastruloid, Cardiogenic Gastruloid and ABV_CDM Culture
2.2.1 Differentiation Medium
Several different media compositions were used during Gastruloid, Cardiogenic
Gastruloid, and Biobot cultures. Basal Media was N2B27 which consisted of equal volumes
10

Neurobasal Medium (GIBCO-21103049) and DMEM/F-12 with GlutaMAX
(GIBCO-313310093) supplemented with 100U/mL Penicillin-Streptomycin, 2mM GlutaMAX
supplement (GIBCO-35050061), 1mM Sodium Pyruvate, 1:100 MEM Non-Essential Amino
Acids, 0.1mM β-mercaptoethanol, 1:100 B27 supplement (50x) serum-free (GIBCO-17504001),
and 1:200 N-2 supplement (GIBCO-17502001). N2B27+++ medium consisted of N2B27 basal
medium supplemented with 30ng/mL bFGF (R&D Systems 233-FB-010/CF), 5ug/mL Human
VEGF (R&D Systems 293-VE-010/CF), and 0.5mM L-Ascorbic Acid (Sigma
Aldrich-255564-100G). Cardiomyocyte Differentiation Medium (CDM) consisted of
StemPro-34 (GIBCO-10639011) supplemented with Stem Pro supplement, 2mM L-Glutamine,
0.5mM L-Ascorbic Acid, 5ng/mL Human VEGF, 10ng/mL bFGF, and 50ng/mL FGF10 (R&D
Systems 345-FG-025/CF).
2.2.2 Differentiation Protocols
Gastruloids and Cardiogenic Gastruloids were generated as previously described by van
den Brink et al., 2014 and Rossi et al., 2021 respectively. In brief, 300 mESCs were plated in
40µL of N2B27 media in 200µL 96-well Clear Round Bottom Ultra-Low Attachment
Microplates (Corning-7007). After a 48 hour aggregation period, a pulse of 150µL of N2B27
containing 3µM Chiron was added to each well. After 72 hours 150µL of the medium was
replaced back to N2B27. The Gastruloid protocol maintained aggregates in N2B27 medium for
the rest of the time course with daily medium changes of 150 µL, the total volume in the wells
was maintained at 190µL. In the Cardiogenic Gastruloid protocol, N2B27 was replaced with
N2B27+++ at 96hr and 120hr. After 144 hours, the medium was then replaced back to N2B27.
After 168 hours, aggregates were transferred from the 96 well plate to low adhesion 6-well
11

plates, with each well containing 10-15 aggregates. From 168hr to 240hr the plates were placed
on an orbital shaker, to prevent adhesion to the plate walls and surface, at 100 rpm in 37℃, 5%
CO2
.
The ABV-CDM protocol begins similarly with the plating of 300 mESCs n 40µL of
N2B27 media in 96-well Clear Round Bottom Ultra-Low Attachment Microplates. After a 48
hour aggregation period, a pulse of 150µL of N2B27 containing 5ng/mL Activin A (R&D
Systems 338-AC-010/CF), 5ng/mL of Human VEGF, and 0.25ng/mL of BMP-4 (Gibco
120-05-5UG). After 72 hours 150µL of the medium was replaced back to N2B27. At 96 hours
150µL of medium was replaced with the CDM media, described above, and again at 120hr. After
144 hours, the medium was then replaced back to N2B27. After 168 hours, aggregates were
transferred from the 96 well plate to low adhesion 6-well plates, with each well containing 10-15
aggregates. From 168hr to 240hr the plates were placed on an orbital shaker at 100 rpm in 37℃,
5% CO2
.
2.3 Live Imaging and Image Analysis
Brightfield and fluorescence imaging were performed daily for each aggregate from 72 hr
to 168hr with Zeiss inverted microscope Axiovert 5 TL at 5x. Images were analyzed using
MOrgAna: Machine-Learning Based Organoids Analysis developed by Gritti et al., 2021.
MOrgAna software was installed locally via python and operated within a Conda environment.
In brief, TIF image files were imported into the software. A training set consisting of 5-8 images
were used to generate a model by manually segmenting their masks. Once the model was trained
it was applied to the batch of images, creating binary masks for each. Morphological
quantifications for area, eccentricity (ratio of the focal distance over the major axis length), and
12

major axis-length were calculated by the software using the generated masks. CSV files of the
data were then exported for visual representation and statistical analysis.
2.4 Calcium Imaging and Dynamics
To visualize calcium flux, aggregates were incubated for 30 minutes with green
fluorescence calcium probe 8µM Fluo-4 (Invitrogen-F10489) at 240 hr in 37℃, 5% CO2
. After
probe incubation medium was replaced with fresh medium and aggregates were incubated at
37℃, 5% CO2 for 10 minutes prior to imaging. Aggregates were then imaged using a Keyence
BZ-X710 equipped with a 37℃, 5% CO2 chamber. The GFP channel was recorded for 100
frames at 200ms. The recorded videos were then analyzed using Image J image software.
Regions of interest (ROI) within the aggregate were drawn around localized foci of contraction
and non-contracting areas using the polygon and ROI manager tools. The area of contraction was
obtained using the measure function in Image J. Fluorescence intensity in each region was
graphed over time. The period of these oscillations were obtained using pyBOAT software
developed by Mönke et al., 2020. In brief, raw oscillatory time series data was imputed into the
software, processed for optimal detrending, and then underwent a wavelet transformation and
ridge analysis to output the period in seconds.
2.5 Motility Assay
To visualize the capacity to move, aggregates in a 6 well low adhesion plate were imaged
using a Keyence BZ-X710 equipped with a 37℃, 5% CO2 chamber with a Nikon 1x objective.
One image was taken every 5 minutes over a thirty minute period to observe motility. Videos
were then analyzed using the Imaris software developed by the Oxford Instruments Group. In
13

brief, the software labeled each individual aggregate as an object and tracked its center of mass
across the video. The software then output values for total track length traveled, total
displacement, average speed, and track straightness for each aggregate which were used to
quantify the motility.
2.6 Motility and Morphological Feature Analysis
Features of contractile length and micro-drift were calculated using Tracker video
analysis and modeling software. Tracker is an open source physics tool built within the Java
framework. Brightfield videos taken during the calcium imaging assay for each individual
aggregate are uploaded into the software. The calibration scale is set to the micron length of the
height and width of the images. Two point masses are created to be tracked, both are placed on
the perimeter of the aggregate. The first point mass, A, is placed on a region that is actively
contracting. The origin of the x,y axis is placed at the location of this point mass. The second
point mass, B, is placed in a region where there is no contractile displacement. Each point is
tracked over the frames of the video. Then, the Tracker software outputs the X and Y coordinates
of the point per frame. These values are then exported to a CSV file. The distance from the start
is calculated using the distance formula:
d = sqrt{(x_2 - x_1)^2 + (y_2-y_1)^2} where x1 and y1 are the coordinate points at frame 0 and
x2 and y2 are the coordinates of that particular frame. The distance from start over time is then
graphed for each point mass individually. Contractile length is calculated by subtracting the
distance from start at the peak of contraction by the distance from start prior to the peak from the
14

graph of point A. The micro-drift is calculated as the final distance from start from point mass B.
Displacement per contraction is calculated by dividing the micro-drift length by the number of
contractions captured in the video. This analysis is performed individually for each aggregate.
2.6 Cadherin Cell Line Generation
In order to generate mESCs that overexpress cadherins we used the ZX1 inducible
cassette system designed by Lacaovino et al., 2011. As previously described, this system entails
designing a plasmid of interest to be inserted as a single copy in the genomic region upstream of
the HPRT housekeeping gene on the X chromosome.
2.6.1 Plasmid Design and Cloning
Four unique cell lines were designed for this project with experimental purpose in mind.
The first cell line was generated to constituently express a fluorescent protein as a tag for wild
type cells in culture. The plasmid design consisted of pairing the CAG promoter, a strong
synthetic promoter to drive gene expression, to GFP, a green fluorescent marker in front of the
LoxP sites. The other three cell lines were generated to have doxycycline inducible
overexpression of cadherins. Cell lines individually overexpressing N-Cadherin, P-Cadherin, and
E-cadherin were generated. Plasmids of these cell lines were designed to constitutively express
mCherry, a red fluorescent marker and have the doxycycline inducible cadherin paired with
iRFP, infrared fluorescent marker, expression. To do so, mCherry was inserted in front of the
PGK promoter followed by T2A, a 2A peptide to induce ribosomal skipping and generate a
polyprotein. Then, following the LoxP sites, the cadherin of interest was linked to an iRFP via
P2A, a different 2A peptide. This allowed the mCherry expression to be constitutively expressed,
15

while the cadherin, tagged with iRFP, was under doxycycline control. Maps for each plasmid can
be seen in Figure 16.
Plasmids were cloned by first conducting polymerase chain reaction (PCR) to amplify the
sequences of interest from donor plasmids. All of the donor plasmids and associated primers are
listed in Table 1. The reaction mixture for the PCR amplification consisted of 12.5µL CloneAmp
HiFi PCR Premix, 2.5µL forward primer, 2.5µL reverse primer, 30 ng of DNA template, di H2O
until total volume of 25µL. Mixtures were then placed into a thermocycler and cycled through
denaturation, 98℃ for 10 seconds, annealing, 55℃ to 65℃ depending on primer for 10 seconds,
extension, 72℃ for 10 seconds, for 30 cycles. Following amplification, purification of the
plasmid was performed on a 1% agarose gel at 110 Volts for 40 minutes. After migration, the cut
plasmid was collected by cutting the appropriate gel band. Next, the DNA was purified from the
gel using the Takara Bio NucleoSpin Gel and PCR Clean-Up kit protocol. Concentrations of
purified DNA were measured with a Nanodrop. Finally, the vectors of interest were assembled
using a Gibson Assembly. The reaction mixture consisted of 10µL NEB HiFi 2x buffer, 4µg of
backbone, 2µg of each vector fragment, and distilled H2O until total volume of 20µL and was
incubated at 50℃ for 1 hour. This process was completed for each of the 4 plasmids.
2.6.2 Bacterial Transformation
Chemically competent (10-beta cells, NEB) bacteria were transformed with the prepared
plasmids. 5µL of Gibson Assembly mix was added to 50µL of bacterial mix and incubated on
ice (4℃) for 30 minutes then at 42℃ for 45 seconds, and then replaced on ice for 5 minutes.
Next, 450µL of SOC medium, prewarmed at 37℃, to the mixture and incubated with shaking,
16

160-225 rpm, for 1 hour at 37℃. 150µL of bacterial mixture was then spread, with beads, on an
ampicillin selective agar plate and incubated at 37℃ overnight.
Following overnight incubation, a confirmation of positive clones was deduced from a
high ratio of colonies in the cloned plate compared to the negative plate. Positive clones were
then individually picked and pipetted into 2mL LB broth with ampicillin. Roughly 6-10 clones
were picked for cultivation per plasmid. The clones were then incubated overnight at 37℃ with
shaking. Following incubation, the DNA extraction was completed using the protocol and
reagents in the ZR Plasmid Miniprep- Classic (Zymo Research-D4015). Eluted DNA was sent to
sequencing via Primordium to confirm correct plasmid construct without mutations.
2.6.3 Transfection into mESCs and Selection
Transfection was performed once the plasmid sequencing results were confirmed to be
positive. The day prior to transfection 1ug/mL of doxycycline was added per well of mESCs to
be transfected. In one tube, a total of 1µg of plasmid was added to 250µL Opti-MEM
(GIBCO-31985062). In a second tube, 250µL Opti-MEM was mixed with 2µL of Lipofectamine
2000 (Invitrogen-11668019). Next, mESCs to be transfected were collected. Collection consisted
of removing medium, rinsing with PBS, detaching with StemPro Accutase, blocking with
DMEM 10% FBS, and centrifugation. Next, prepared Opti-MEM and Lipofectamine 2000 were
mixed with plasmids and left to incubate for 10 minutes. mESCs were resuspended in 1mL of
PBS and counted. 200,000 cells were aliquoted per sample and mixed into tubes with
Opti-MEM, Lipofectamine 2000, and plasmid and incubated for 10 minutes. Following
incubation, solutions were diluted with 5mL of medium, centrifuged, and replated. The following
17

day 300 ug/mL Neomycin (Thermo Scientific Chemicals-J67011-AD) was added to the medium.
Selection took place over the next 10 days with daily medium change.
After the Neomycin selection period, cells were collected in FACS buffer (1/4000
Hoechst) for sorting. GFP+ and mCherry+ cells, for respective cell lines, were sorted by FACS
(Aria II, BD Biosciences) and replated and expanded in ES maintenance medium.
2.7 Doxycycline Time Course Experiment
In order to test the effect of prolonged doxycycline exposure on the system a time course
experiment was conducted. In a 24 well cell culture plate (Falcon-353047) 50,000 cells were
seeded per well in 5 Gelatin coated wells in ESC maintenance media. At 0hr, 0.5 ug/mL of
doxycycline was added to one of the wells. Every consecutive day 0.5 ug/mL doxycycline was
added into one new well. Fresh ESM maintenance media was placed into every well, and those
that had doxycycline already added received fresh 0.5 ug/mL. At 96 hours every well had
received doxycycline for a varied amount and one well had not received it as a negative control.
The cells were collected and mCherry+ cells were counted via FACS (Attune NxT). This
experiment was conducted for all 3 of the cadherin overexpressing cell lines.
2.8 Flow Cytometry Analysis
Flow cytometry was performed on a variety of samples: 3D aggregates at various time
points along the differentiation protocol and 2D cultures of generated cell lines. Several
antibodies were used to quantify the expression of different proteins and all of the antibodies
used can be found in Table 2. Samples were first collected into a 15mL Falcon tube and rinsed
with 2mL of PBS. Samples were then incubated in 1mL of 3mM Ethylenediaminetetraacetic acid
18

(EDTA) (ThermoScientific Chemicals-J15694.AP) for 10 minutes in a 37℃ water bath. After
incubation, samples were mechanically dissociated by pipetting up and down with a P200 set at
200µL. Dissociated samples were then neutralized with 2mL of DMEM and centrifuged for 4
minutes at 400 rpm. For every step onwards, samples were kept on ice and the centrifugation was
done in the cold room at 4℃. Following centrifugation, samples were resuspended in 1mL of
FACS buffer and transferred to a 1.5mL Eppendorf tube. Samples were then centrifuged again
and resuspended in 1mL of FACS buffer.
For cell lines without fluorescent reporter tags, samples were stained with appropriate
antibodies. First, antibodies were diluted, respectively to manufacture instructions, and added to
samples. Tubes were placed on ice for 30 minutes with foil covering. After incubation, 1mL of
FACS buffer was added to the samples and centrifuged for 4 minutes at 400 rpm. Next, samples
were resuspended in 500µL of FACS buffer with a 1:2000 Hoechst dilution. Samples were then
filtered through the FACS tube and placed on ice. The antibodies used in this thesis are listed in
Table 1.
2.9 Immunofluorescence
Immunofluorescence was performed on a variety of samples: 3D aggregates at various
time points along the differentiation protocol and 2D cultures of generated cell lines. Therefore,
fixing and incubation steps were optimized for each to ensure proper antibody penetration and
binding. To start, samples were fixed with 4% paraformaldehyde PFA for 30 minutes at room
temperature, thick aggregates were placed on a rocker at 50 rpm and incubated for 1 hour in the
cold room at 4℃. Next, samples were rinsed 3 times with PBS, once instantaneously and twice
with a 10 minute incubation period in between. Samples were then suspended in a blocking
19

solution consisting of 1% PBS-BSA, 0.2% Triton, and 5% Donkey Serum for 1 hour on rocker at
50 rpm. After incubation, primary antibody was added into the blocking solution and left
overnight in the cold room. The full list of antibodies is located in Table 3. The next morning the
samples were rinsed 3 times with PBS as mentioned previously. The secondary antibody was
then added to the sample in blocking solution and 1/2000 Hoechst and incubated at room
temperature for 1-2 hours on the rocker at 50 rpm. Finally, the samples were rinsed 3 times with
PBS as mentioned previously. Samples were then mounted onto slides, PBS was used as the
mounting medium, and left to dry overnight in the cold room. Imaging was performed using
Zeiss Axiovert 5 TL and Keyence BZ-X710. The antibodies used in this thesis are listed in Table
1.
2.10 Statistical Analysis
Graph Pad Prism 9.0 was used to generate graphs and perform statistical analysis.
One-way analysis of variance (ANOVA) tests followed by Bonferroni’s test were used to
compare means across 2 or more groups. Data shown in column graphs are expressed at mean ±
standard deviation.
20

Chapter 3: Results
3.1 Specific Aim 1: Optimization of a protocol for the generation of mESC-derived
elongated and contractile cellular constructs via a soluble factor approach.
3.1.1 Capacity for current organoid protocols to generate contractile aggregates
To begin, we sought to explore the ability of current morphogen-based protocols for
derivation of organoids with contractile cells that could generate aggregates with the capacity for
motility. Gastruloids and cardiogenic gastruloids were generated as previously described by van
den Brink et al., 2014 and Rossi et al., 2021 respectively. In brief, 300 mESCs-E14 were plated
in 40µL of N2B27 media in 96-well Clear Round Bottom Ultra-Low Attachment Microplates
(Corning-7007). After a 48 hour aggregation period, mesoderm induction was conducted via a
pulse of CHIR for 24hr. Gastruloids were maintained in the basal N2B27 medium for the
remaining duration of the differentiation. Cardiogenic gastruloids received a second induction
from 96hr to 144hr with a cocktail of cardiogenic factors consisting of ascorbic acid, bFGF, and
VEGF (Fig. 4).
We observed that the gastruloid protocol had a very low percentage of spontaneously
beating aggregates at day 7. 39.51±6.95% of the aggregates were contracting with the
cardiogenic gastruloid protocol, consistent with similar efficiency reported by Rossi et al (Fig.
7).
3.1.2 Effect of ABV_CDM on aggregate area, eccentricity, and length
In order to increase the efficiency of contractile aggregate generation we optimized the
protocol to induce cardiomyocyte differentiation. To do so we set up a differentiation protocol
21

with the same induction windows as those described in Rossi et al., 2021, but with alterations to
the factors delivered. For the first induction at 48 hr we replaced the pulse of CHIR with a pulse
of Activin A, BMP4, and VEGF. Manipulation of these specific cardiogenic factors has been
shown to be vital for cardiomyocyte differentiation. Activin A supports mesoderm and endoderm
induction, BMP4 patterns the mesoderm, and VEGF supports angiogenesis. For the second
induction window from 96 hr to 144 hr we replaced the N2B27 +++ media with CDM, the
contents of which is described previously. Notably, CDM consists of ascorbic acid, bFGF,
FGF10, and VEGF (Fig. 5) We then sought to compare the morphology of aggregates generated
by these alterations. To do so, we took bright field images at 24 hour intervals starting from 96 hr
(Day 4) to 168 hr (Day 7). Images were then analyzed using the MOrgAna software developed
by Gritti et al. 2021. First, we looked at the major axis length. We found that on average
aggregates from the ABV_CDM protocol had shorter lengths at each time point and in their final
form, although the difference was not significant. The measured mean major axis lengths on day
7 were: 1131±121µm, 1260±143µm, and 1019±101µm for gastruloids, cardiogenic gastruloids,
and ABV_CDM aggregates respectively. Furthermore, we observed that aggregates from both
the gastruloid and cardiogenic gastruloid protocol experienced a great elongation event between
96 hr to 120 hr consistent with their reported papers, but ABV_CDM aggregates did not (Fig
6A). We also looked at the area of the aggregates. There was a similar trend as major axis length,
gastruloids and cardiogenic gastruloids had a greater increase in area from 96 hr to 120 hr. The
measured mean areas on day 7 were: 408,373±80,104µm2
, 426,756±73,312µm2
, and
383,230±68,115µm2
for gastruloids, cardiogenic gastruloids, and ABV_CDM aggregates
respectively (Fig. 6B). Finally, we looked at eccentricity. Eccentricity is the ratio of the focal
distance over the major axis length. The value approaches 0 as the shape becomes more circular.
22

The general trend was an increase in eccentricity as the aggregates morphed into their elongated
form. At 120 hr ABV_CDM aggregates showed lower levels of eccentricity compared to the
other two protocols, again highlighting a difference between 96 hr to 120 hr. The measured mean
eccentricities on day 7 were: 0.655±0.128, 0.833±0.075, and 0.687±0.121 for gastruloids,
cardiogenic gastruloids, and ABV_CDM aggregates respectively (Fig. 6C).
3.1.3 Effect of ABV_CDM on efficiency of deriving contractile aggregates on Day 6 and Day 7
We next looked at the efficiency of the ABV_CDM protocol to generate contractile
aggregates. We found that 86.31±7.6% of aggregates were contracting by day 7 with the
ABV_CDM protocol. Furthermore, we saw that 36.83±6.05% of aggregates were spontaneously
contracting by day 6 (Fig. 7). This was before any gastruloids or cardiogenic gastruloids were
observed to be contracting.
3.1.4 Effect of ABV_CDM on percentage area of Contraction and calcium dynamics
We next sought to characterize the contractility observed in these aggregates. To do so,
we incubated aggregates with a Fluo-4 calcium probe and recorded the calcium activity (Fig
8A-C). We next determined the percent area of contraction by measuring the region with
fluorescence change using the polygon tool on FIJI (Fig. 8D). We found that ABV_CDM
aggregates had a mean percentage area of contraction equal to 54.10±11.62%, compared to a
mean percentage area of contraction equal to 21.08±9.55% for contracting cardiogenic
gastruloids. Next, we determined the period of contractions from the change in fluorescence
intensity over time using the pyBOAT software developed by Mönke et al., 2020 (Fig 8E). We
found that ABV_CDM contractile aggregates had higher frequencies of contraction compared to
23

contracting cardiogenic aggregates (Fig. 8G). ABV_CDM aggregates had a frequency of
0.98±0.13Hz and cardiogenic gastruloids had a frequency of 0.45±0.08Hz.
3.2 Specific Aim 2: Characterization of the developmental trajectory and emergent
function of 3D cellular constructs.
3.2.1 Effect of ABV_CDM on Flk1 and PDGRFα cardiac mesoderm markers
In order to determine how these motile aggregates were generating more cardiomyocytes,
we sought to understand their developmental trajectory. First, I looked at the quantity of cells
induced to be cardiac mesoderm with cardiac potential. To do so, I conducted FACS analysis for
Flk1 and PDGRFα. Flk1 is a known marker for mesoderm. Double positive Flk1+ and PDGRFα+
marks induced cardiac mesoderm with cardiac potential. These markers are transient and peak
earlier along the differentiation time course around day 4. I collected aggregates daily from day 2
to day 7 along the differentiation and analyzed them for FACS. I hypothesized that the
ABV_CDM protocol would generate a greater Flk1+PDGRFα+ population at this time point.
However, we found that there was no significant difference in the cardiac mesoderm population
across all the protocols, even the standard gastruloid protocol. All protocols had peak expression
of 90% at day 4 and a sharp decline over the next 48 hours (Fig 9). No major conclusions can be
drawn as the data reported is only n=1.
3.2.2 Effect of ABV_CDM on Nkx2.5 expression
I next looked at the expression of Nkx2.5 across each of the protocols. Nkx2.5 is a known
marker for cardiac progenitors and plays a vital role in cardiac development. This marker is
24

expressed later in the cardiac lineage trajectory around day 6. To quantify its expression we used
the Nkx2.5-eGFP reporter cell line from Hsia et al., 2008. I collected aggregates from each
protocol on Day 4, 5, 6, 7, and 10 and analyzed them via FACS. I found that the ABV_CDM and
Rossi et al., 2021 protocols had increased Nkx2.5 population at day 6 and day 7, which aligned
with the expected developmental window. ABV_CDM had a significant increase in population
compared to the Rossi et al., 2021 protocol with a peak of 36.54±3.82%. Rossi had a peak
expression of roughly 12.73±1.14% (Fig 10).
3.2.3 Effect of ABV_CDM on Brachyury and Sox-17 expression
We next sought to study the induction of mes-endoderm and endoderm across the
protocols. Endoderm derivatives have been shown to play an important role in supporting cardiac
differentiation. To quantify its presence I used Sox-17-RFP,Bra-GFP cell line from Pour et al.
2022. I collected aggregates from each protocol on day 4, 5, 6, 7, and 10 and analyzed them via
FACS. I found that peak levels of Brachyury were similar among all protocols, approximately
37.53%. However, there was an apparent delay in Brachyury suppression in the ABV_CDM
protocol, with higher levels at day 5 compared to the other two protocols (Fig. 11A). Similarly,
there was no significant variation in Sox-17 expression across protocols. All protocols peaked
around day 6 around 21.24% (Fig. 11B). The paired dynamics of all expression markers for the
ABV_CDM protocol is visualized in Figure 12.
3.2.4 Effect of ABV_CDM on germ layer distribution at Day 5
To further investigate the distribution of germ layers within the aggregates Dr. Fokion
Glykofrydis, a postdoctoral scholar in the Morsut Lab, conducted IF stained samples I fixed on
25

Day 5 of the differentiation protocol. We then analyzed the germ layer distribution by looking at
the following markers: DAPI, Sox2, Sox17, and Eomes. The expression of Sox2 is a known
marker for ectoderm, Sox17 is a known marker for endoderm, and the expression of Eomes is a
marker for mesoderm. It is known that gastruloids have the ability to give rise to all 3 of these
germ layers, but we sought to investigate their spatial distribution across protocols. We found
that in standard gastruloids and cardiogenic gastruloids Sox2 and Sox17 were polarized at the
anterior axis by Day 5, consistent with published studies. Eomes was polarized, but more evenly
distributed across the aggregate. We found that ABV_CDM had similar polarization of these
markers at Day 5. Interestingly, Sox17 seemed to be more widely distributed in the ABV_CDM
aggregates (Fig. 13). However, these observations are purely qualitative and deeper quantitative
analysis of area overlays and intensity distribution need to be conducted.
3.2.5 Effect of ABV_CDM on emergence of motility at Day 10
On Day 10 of the differentiation protocol we assessed if aggregates had the capacity to
move. To do so, Dr. Christine Ho, a postdoctoral scholar in the Morsut lab, and I set up a motility
assay using the Keyence BZ-X710 equipped with a 37℃, 5% CO2 chamber with a Nikon 1x
objective. Aggregates were placed in batches of 10-15 per well in a 6 well low adhesion plate.
Aggregates were then recorded for thirty minutes with images taken every 5 minutes. The videos
were then analyzed using the Imaris image analysis software to quantify the motility of the
aggregates. An image of the Imaris software output is shown in Figure 14A. Aggregates for each
protocol were then pooled together and their total displacement. We found that aggregates
generated from ABV_CDM add the greatest displacement across the three protocols, with an
26

average displacement of 788.67±289.56µm in 30 minutes (Fig. 14B). The average speed was
0.26±0.01µm/s for ABV_CDM aggregates (Fig. 14C).
3.2.6 Associating morphological and functional features with motility
We next sought to determine associations between morphological and functional features
with motility observed during the batch experiments mentioned above. To do so, I devised
metrics to characterize the morphotype of individual aggregates. For morphological features, we
calculated the total area, aspect ratio, and circularity of the adult form at Day 10. These features
were extracted through image analysis using Image J. For functional features I quantified the
contractile length of the aggregate, drift, and displacement per contraction. These features were
calculated using the Tracker Physics Analysis software. Contractile length was defined as the
distance between contracted and relaxed state of the aggregate. Lengths were plotted on a
Distance From Origin vs Time graph. Drift was defined as the displacement of the entire
aggregate. Displacement per contraction was calculated by dividing the drift by the number of
contractions recorded. I also paired the calcium period and calcium area percentage of each
aggregate to these features. I aimed to correlate these metrics to the displacement, average speed,
and track straightness of the individual aggregates. The paired characteristics for each individual
aggregate are shown in Figure 15. Based on these initial analyses I hypothesize that a greater
contractile length and contractile percentage, correlates to faster motility and greater distance
traveled. However, these are very preliminary hypotheses based on current analysis. A deeper
principal component analysis needs to be conducted with more samples that are moving and not
moving.
27

3.3 Specific Aim 3: Utilization of the ZX1 ICE system to synthetically engineer control of
tissue patterning and body plan.
3.3.1 Design of cadherin overexpressing cell lines
In order to control the spatial body plan of the tissue we generated cell lines that
conditionally overexpress a unique cadherin. To do so we used the ZX1 ICE system. This system
allows for the doxycycline inducible overexpression of a gene of interest. I generated three total
cell lines, one for N-Cadherin, E-Cadherin, and P-Cadherin respectively. Plasmids of these cell
lines were designed to constitutively express mCherry and conditionally overexpress the
cadherin of interest and iRFP. A fourth cell line was generated that constitutively overexpressed
GFP as a proxy to track wild-type cells in co-culture experiments. The plasmid designs are
shown in Figure 16.
3.3.2 FACS and immunofluorescence analysis for cadherin overexpression after 24 hour
doxycycline induction
In order to confirm the overexpression of each cadherin I conducted FACS and
immunofluorescence staining. Cells were incubated with 0.5 ug/mL doxycycline for 24 hours
and then collected for analysis. Engineered fibroblast L929 cell lines, from the Morsut Lab,
known to overexpress cadherins upon dox induction were used as positive control for the
N-Cadherin and P-Cadherin antibodies. I found that there was no appreciable increase in
cadherin expression upon doxycycline induction for any of the cell lines. N-Cadherin had a
non-significant increase (Fig. 17). These results were further confirmed with
28

immunofluorescence for each cadherin. There was no quantifiable increase in cadherin antibody
staining before and after doxycycline induction (Fig. 18-20).
3.3.3 Effect of prolonged doxycycline induction response on mCherry signal
To further characterize the generated cell lines I conducted a doxycycline induction time
course experiment. Cells were induced with doxycycline for 0hr, 24hr, 48hr, 72hr, and 96hr and
then collected for FACS analysis of mCherry expression. I found that longer doxycycline
induction led to a decrease in mCherry expression for each cell line (Fig. 21).
29

Chapter 4: Discussion
Tissue engineering and biorobotics propel our understanding of cellular coordination,
multicellular assembly, and the interplay between form and function. Efforts have probed the
capacity to use cellular components as actuators, sensors, and structural frames, integrating them
with non-living material to create bio-hybrid machines. Recently, efforts have explored the
ability to engineer cellular-only constructs of biorobots. This work aims to produce novel
anatomic configurations and emergent functions from the intrinsic machinery of cells. Thus,
pushing our knowledge of cell plasticity and collective behavior. Several successful attempts
have been made in this pursuit. Most recently, Gumuskaya et al., 2023 published a protocol for
Anthrobots. These are self-organizing and motile biobots of stem cells derived from adult human
lung tissue. Anthrobots are cultured in an extracellular matrix for two weeks and their final
locomotion property is powered by cilia.
This thesis puts forth a new player in the realm of biobots and biological engineering: a
self-motile cellular construct via cardiomyocyte actuation. Unlike previous constructs, this
protocol entails the use of mouse embryonic stem cells that are coaxed towards a novel body
plan and myocyte powered motility. The protocol is 10 days long, with emergency of
contractility and motility by days 6 and 7 respectively, and consists of soluble factor delivery
during specific temporal windows. The final constructs are a mean 1019±101µm in length. In a
motility assay on day 10 they display a mean displacement of 788.67±289.56µm.
4.1 Impact of temporal delivery of soluble factors on tissue developmental trajectories
One strategy employed in traditional tissue differentiation protocols is the delivery of
soluble factors in temporal windows that aim to mimic those during development in vivo. For
30

cardiac differentiation specifically the efficiency of a stage-specific differentiation protocol with
dynamic delivery of Activin A and BMP4 factors was highlighted by Kattman et al., 2011. This
strategy continued into protocols for 3D organoids with a 24 hour Chiron pulse at day 2 to
initiate gastrulation in Gastruloids and a 48 hour delivery of three cardiogenic factors from day 4
to 6 in Cardiogenic Gastruloids (van Den Brink et al., 2014; Rossi et al., 2021). In order to
explore the possibility of developing a de novo tissue structure and function we employed the
strategy of non-canonical delivery of cardiogenic factors.
The protocol presented in this thesis consists of two main alterations to traditional
gastruloid and cardiogenesis protocols. The first is the replacement of a Chiron pulse with a
pulse of Activin A, BMP4, and VEGF at day 2. The next is the delivery of cardiomyocyte
differentiation medium which consists of Ascorbic acid, VEGF, bFGF, FGF10, and a stem pro
base in a second induction from day 4 to 6. These adjustments led to different not only a novel
function and tissue, but also a varied developmental trajectory. Geometrical and size differences
were seen in the aggregates derived from this protocol. Notably, there was a delayed elongation.
Traditional aggregates with a Chiron pulse showed elongation and symmetry breaking from day
4 to 5 while the ABV pulse aggregates did not show marked growth until day 5 to day 6. The
exact reason for this is unknown; however, one postulation is that without the exogenous Chiron
the cells lack the impetus to migrate and extend the anterior-posterior axis. A further hypothesis
is that the lack of elongation and maintained roundness of the aggregates allowed for a geometry
conducive to cardiomyocyte differentiation.
Molecular characterization of the ABV_CDM protocol led to a mix of expected and
unexpected results. To begin, there was a great increase in Nkx2.5+, a late-stage marker for
cardiac progenitor cells, in the ABV_CDM protocol. This was expected as these aggregates had
31

higher percentages of contractile area and higher frequencies of contraction. Contrarily, it was
surprising that there were no significant differences in Flk+PDGRFa+, a double positive
population for mesoderm populations with cardiogenic potential. It is still yet to be determined
how a similar pool of potential cardiomyocytes are actualized as differentiated myocytes within
this protocol. The underlying molecular mechanisms resulting from altered factor delivery, and
that ultimately govern the aggregate’s collective behavior are important to uncover. Further
analysis will need to focus on the factors influencing the cell fate decision during this
developmental trajectory. One potential indicator is the spatial distribution of these cell markers.
Future work can utilize reporter cell lines to map the live dynamics of their distribution in the
aggregates. Overall, this work contributes to the notion that biological constructs with de novo
forms and functions can be built within a cell’s natural genome as long as sufficient and timely
growth factor stimulation is provided.
4.2 Morphological properties associated with motility
Preliminary work has been done on associations between morphology and the emergent
function of motility in our ABV_CDM aggregates. Several features of the aggregates were
collected and analyzed for correlations between motility. Morphological features considered
were area, length, and eccentricity. Contractile features considered were contractile area
percentage, polarity, frequency, and contractile length. Motility features considered were average
speed, total displacement, and track straightness. Our early analysis indicates that increased
contractile length, area % of contraction, and frequency are important for motility. Increased
numbers of aggregates analyzed will hopefully elucidate distinct morphotypes.
32

Limitations exist in the presented analytical methods. Foremost, all of the analysis has
been done solely on 2D images or videos of the aggregates. Future analysis should include
crucial 3D structural characteristics using techniques such as confocal imaging. This will help
elucidate the physics behind the motility. The cardiomyocytes are generating a contractile force;
however, measurement of the magnitude of this force and the physical deformations that lead to
movement are important to dissect. Furthermore, the ratio of contractile area to noncontractile,
“payload”, area is postulated to play a role in aggregate speed and total displacement. This
payload is yet to be thoroughly characterized. Confocal sectioning will provide information on
the cellular composition of the aggregates that make up the entire tissue architecture. With the
deciphering of these characteristics, the long term vision of being able to enlist certain protocols
to output specific motility types, and other emergent functions, can be realized. Fine-tuned
developmental programs can be designed to make aggregates with user-defined functions.
4.3 Challenges of overexpressing cadherins in ECSs
The attempt to overexpress cadherins in ESCs was unsuccessful in this thesis as seen in
Figures 17-20. The original goal of generating the cell lines was to drive phase separation, via
differential cadherin adhesion, in order to exert control over the tissue body plan. Each cell line
was poised to adhere homogeneously, thereby creating compartmentalization within the
aggregate. However, the cell lines produced did not overexpress cadherins to a sufficient level in
order to drive this homotopic separation. Several postulations may explain why this specific
attempt was unsuccessful.
First, the chosen ZX1 ICE system may not have been best suited for this goal. While the
system has advantages in efficient recombination and reduced off target mutations, it lacks
33

strength in its ability to overexpress genes due to single copy integration. Any gene of interest is
only integrated as a single copy at the HPRT locus. Although this locus is suitable for transgene
expression, a system with multicopy integration may be better suited in order to generate
significant overexpression of cadherins in ESCs. Furthermore, the chosen PGK promoter may
not be strong enough to initiate transcription. A stronger promoter such as EF-1ɑ or CBA could
enhance transcription and ultimately cadherin expression (Norrman et al., 2010).
Second, the design of the system and size of the cadherin gene may have impacted the
ability of the cell lines to overexpress the transgene. As seen in Figure 21, there was a slight
decrease in mCherry expression due to prolonged doxycycline induction. This was an
unexpected result as mCherry was designed for constitutive expression. One proposed reason for
this finding is that the chosen T2A linking system is not sufficiently supporting protein
translation upon doxycycline induction. The cadherin genes themselves were approximately
2,500 base pairs in length. The great size of these proteins may inhibit their efficient translation
and be beyond the threshold for the ZX1 system. Furthermore, the design of the plasmids paired
each cadherin with the iRFP protein via P2A, adding another level of complexity to the system.
Although no mutations were noted in the sequencing, a reduction in design may have been
beneficial for overexpression.
Finally, if we suppose that the ZX1 system was effective in translating the transgene, the
lack of overexpression may be due to endogenous control mechanisms within ESCs. Cadherins
are pertinent genes during differentiation and are dynamically expressed. For example, during
embryogenesis cadherins are tightly regulated and the transition between different cadherins is
vital for cell sorting, movement, and polarization (Soncin and Ward 2011). Furthermore,
differential cadherin expression is seen in cell fate specification. The down regulation of
34

E-Cadherin and upregulation of N-Cadherin has been documented in both cardiac and neural
differentiation (Punovuori et al., 2019; Yan et al., 2020). Given their role in cell identity,
stringent molecular mechanisms govern cadherin expression. Therefore, it can be hypothesized
that even if the transgene was successfully overexpressed, endogenous mechanisms could
downregulate it. This work highlights key considerations and difficulties to address in future
engineering attempts.
35

Chapter 5: Tables and Figures
Figure 1. Current models of myocyte powered biohybrid robots. (adapted from Nawroth et
al., 2012; Young Lee et al., 2022; Yuan et al., 2023). (A) Design and fabrication of the Medusoid,
a silicone polymer scaffold with strategically placed rat cardiomyocytes to mimic the natural
swimming dynamics of jellyfish. (B) Design and assembly of biohybrid fish, a gelatin construct
with a muscle bilayer of human stem cell derived cardiomyocytes, that can be optically induced
to have mechano-electrical responses conducive to swimming via contraction. (C) Overview of
constructs that harness the contractile force of myocytes for engineered purposes.
36

Figure 2. Schematic of Activin/Nodal and BMP Signaling. (adapted from Casanova et al.,
2011). Activin/Nodal and Bone Morphogenetic Protein-4 pathways are both part of the TGF-β
family of morphogens. Activin and Nodal both act on transmembrane Activin Type 1 and II
serine/threonine kinase receptors, which phosphorylate Smad2 and Smad3. These complex with
Smad4 and translocate to the nucleus to regulate transcription. BMP binds to BMP Type 1 and II
serine.threonine kinase receptors, which phosphorylate Smad1, 5, and 8. These complex with
Smad4 and translocate to the nucleus to control transcription. Both of these pathways are shown
to be implicated in cardiac differentiation from embryonic stem cells.
37

Figure 3. Schematic of ZX1 Inducible Cassette Exchange and Tetracycline-On System.
(adapted from Lacovino et al., 2011; ). (A) The ZX1 ICE system utilizes LoxP homologous
recombination to insert a transgene of interest upstream of the HPRT locus under conditional
control of the TRE promoter. (B) In the absence of doxycycline rtTA remains unbound to TRE
and transcription is repressed. Doxycycline binds to rtTA and the complex binds to TRE to
initiate target gene transcription. Thus, the Tet-On system allows for conditional control of
transgene expression.
38

Figure 4. Currently available protocols for 3D Gastruloids and Cardiogenic Gastruloids.
(A and B adapted from van den Brink and van Oudenaarden 2021). (A) Schematic overview of
workflow to generate gastruloids from Day 0 to Day 7, with corresponding embryonic stages.
(B) Schematic describing the cellular composition and compartmentalization of a Cardiogenic
gastruloid at 168hr developed by Rossi et al., 2020). (C) Bright field images of the Gastruloid
protocol developed by van den Brink et al., 2014. A single aggregate was imaged every 24 hours
from Day 4 to Day 7. (D) Bright field images of the Cardiogenic Gastruloid protocol developed
by Rossi et al., 2021. A single aggregate was imaged every 24 hours from Day 4 to Day 7. Scale
Bars equal 500µm.
39

Figure 5. ABV_CDM differentiation protocol for motile aggregates. (A) Brightfield images
of ABV_CDM protocol. A single aggregate was imaged every 24 hours from Day 4 to Day 7.
(B) Table comparing the differences in factors added to culture media during induction windows.
Notably, ABV_CDM replaces the Chiron pulse in the first induction with a pulse of cardiogenic
factors (Activin A, BMP-4, and VEGF). Further, the second pulse consists of a cardiomyocyte
differentiation medium consisting of Ascorbic Acid, VEGF, FGF10, bFGF, and a StemPro base,
replacing the N2B27+++ media used in the cardiogenic protocol. Scale Bars equal 500µm.
40

Figure 6. Morphological features of aggregates generated across protocols. (A) Bar graph
depicting the major axis length in microns of aggregates. (B) Bar graph depicting the area in
microns squared of aggregates. (C) Bar graph depicting eccentricity of aggregates. All
measurements were calculated from 5x brightfield images with MOrgAna software (Gritti et al.,
2021). Results displayed are for n=6, with 3 experiments done by myself and 3 done by Dr.
Christine Ho. Each experiment consisted of 96 aggregates per protocol.
41

Figure 7. Percent of contractile aggregates on day 6 and day 7 of differentiation. Bar graph
depicting the percentage of observed aggregates to be beating. Aggregates were qualitatively
assessed for spontaneous contraction under Zeiss Axiovert 5 TL microscope. Results displayed
are for n=6, with 3 experiments done by myself and 3 done by Dr. Christine Ho. Each
experiment consisted of 96 aggregates per protocol.
42

Figure 8. Calcium dynamics and contractile area percentages of aggregates. (A-C)
Fluorescent images of aggregates stained with Flou-4 calcium probe. Green fluorescence
represents active calcium channels and is a proxy for contractility. (D) Tracing of aggregate into
contractile and noncontractile regions for fluorescent analysis. (E) Fluorescence intensity over
43

time output from pyBOAT software (Mönke et al., 2020) for each region of interest. (F) Bar
graph depicting the average calcium area percentage across protocols. (G) Bar graph depicting
average frequency across protocols. n=6 for bar graphs shown. Scale bars = 500
Scale bars equal 500µm.
44

Figure 9. FACS analysis of Flk1+PDGRFα+ double positive populations across protocols.
Line graph depicting the double positive population for Flk1+PDGRFα+ across protocols via
FACS analysis. Aggregates were collected and analyzed every 24 hours from day 2 to day 7. The
data reported is only equal to n=1.
45

Figure 10. FACS analysis of Nkx2.5+ positive population across protocols. Line graph
depicting the positive population for Nkx2.5+ across protocols via FACS analysis. The
Nkx2.5-eGFP reporter cell line from Hsia et al., 2008 was used for differentiation and analysis.
Aggregates were collected and analyzed every 24 hours from day 4 to day 7. The data reported is
only equal to n=2. Kyle Poon contributed to sample collection and analysis.
46

Figure 11. FACS analysis of Brachyury+ and Sox-17+ positive populations across protocols.
(A) Line graph depicting the positive population for Brachyury+ across protocols via FACS
analysis. (B) Line graph depicting the positive population for Sox-17+ across protocols via FACS
analysis. The Sox-17-RFP,Bra-GFP cell line from Pour et al. 2022 was used for differentiation
and analysis. Aggregates were collected and analyzed every 24 hours from day 4 to day 7. The
data reported is only equal to n=2. Kyle Poon contributed to sample collection and analysis.
47

Figure 12. Paired expression of cell markers during ABV_CDM differentiation protocol.
Line graph depicting the developmental dynamics from day 2 to day 10 for ABV_CDM
aggregates.
48

Figure 13. Immunofluorescence staining for germ layers on Day 5. Aggregates were fixed
and stained on day 5 of the differentiation protocol. Samples were stained for the following
markers: DAPI, Sox2, Sox17, and Eomes. Dr. Fokion Glykofrydis stained samples that I fixed
from culture. We analyzed the images together. Images were taken at 10x on the Zeiss Axiovert 5
TL microscope. Scale bars equal 100µm.
49

Figure 14. Batch motility assay and analysis on Day 10. (A) Sample output from Imaris image
analysis software. Each aggregate’s center of mass is tracked across motility assay. (B) Bar graph
depicting average displacement length at 15 minutes and 30 minutes during motility assay. (C)
Bar graph depicting the average speed of aggregates throughout the motility assay. Dr. Christine
Ho contributed to assay analysis. Scale bar equal to 500µm.
50

Figure 15. Morphological, contractile, and motility feature analysis at individual aggregate
level. (A) Brightfield image of aggregate on day 10. (B) Morphology characteristics analyzed for
individual aggregate. (C) Outlined is the region of fluorescence activity during calcium probe
incubation, indicating contractile area. (D) Aggregate features derived from calcium dynamics.
(E) Image depicting TrackR software point analysis for contractile length and micro motility. (F)
Motility features from motility assay. (G) Line graph plotting the distance from origin over time
of the contracting region. (H) Sample lengths calculated per recorded contraction. (I) Total drift
calculated by finding the final distance from origin. Scale bars equal to 500µm.
51

Figure 16. Plasmid design for generated cell lines with ZX1 inducible cassette exchange
system. (A) Plasmid map for constitutive GFP cells. (C-D) Plasmid maps for cell lines that
conditionally overexpress unique cadherin, Ecad, Ncad, and Pcad respectively. (E) Zoom in on
circuit logic for constitutive mCherry expression and conditional cadherin expression tagged
with iRFP.
52

Figure 17. FACS analysis for cadherin expression after 24 hour doxycycline induction. (A)
Histogram depicting N-Cadherin positive stained population. (B) Histogram depicting
E-Cadherin positive stained population. (C) Histogram depicting P-Cadherin positive stained
population. No appreciable increase is observed from no doxycycline and doxycycline conditions
indicating failed attempt to overexpress cadherin genes.
53

Figure 18. Immunofluorescence staining for N-Cadherin overexpression after 24 hour
doxycycline induction. (A) Negative control without cadherin antibody. (B) Induced cells with
N-cadherin antibody. (C) Antibody staining in the generated CAG-GFP cell line. (D) Positive
control with in house fibroblast L929 cell line. No indication of increased fluorescence with
doxycycline induction. Images taken on Keyence BZ-X710 at 10x Scale bars equal to 50µm.
54

Figure 19. Immunofluorescence staining for P-Cadherin overexpression after 24 hour
doxycycline induction. (A) Negative control without cadherin antibody. (B) Induced cells with
P-cadherin antibody. (C) Antibody staining in the generated CAG-GFP cell line. (D) Positive
control with in house fibroblast L929 cell line. No indication of increased fluorescence with
doxycycline induction. Images taken on Keyence BZ-X710 at 10x Scale bars equal to 50µm.
55

Figure 20. Immunofluorescence staining for E-Cadherin overexpression after 24 hour
doxycycline induction. (A) Negative control without cadherin antibody. (B) Induced cells with
E-cadherin antibody. (C) Antibody staining in the generated CAG-GFP cell line. No indication
of increased fluorescence with doxycycline induction. Images taken on Keyence BZ-X710 at 10x
Scale bars equal to 50µm.
56

Figure 21. FACS analysis for mCherry signal with increasing duration of doxycycline
induction. (A-C) Histograms displaying the mCherry positive population following prolonged
doxycycline induction.
57

Table 1. Antibodies Table of antibodies used for immunofluorescence and FACS analysis in this
thesis.
58
Antibody Catalog Number
Flk1-Biotin Avas12a1, #13-5821-82, Ebioscience
Pdgfra-PE APA5, #12-1401-81, Ebioscience
E-Cadherin 24E10, Cell Signaling Technology
N-Cadherin 610920, BD Biosciences
P-Cadherin 74545, Santa Cruz

References
Blackiston, D., Kriegman, S., Bongard, J., & Levin, M. (2023). Biological Robots: Perspectives
on an Emerging Interdisciplinary Field. Soft Robotics, 10(4), 674–686.
https://doi.org/10.1089/soro.2022.0142
Boch, J., Scholze, H., Schornack, S., Landgraf, A., Hahn, S., Kay, S., Lahaye, T., Nickstadt, A.,
& Bonas, U. (2009). Breaking the Code of DNA Binding Specificity of TAL-Type III
Effectors. Science (American Association for the Advancement of Science), 326(5959),
1509–1512. https://doi.org/10.1126/science.1178811
Bronson, S. K., Plaehn, E. G., Kluckman, K. D., Hagaman, J. R., Maeda, N., & Smithies, O.
(1996). Single-Copy Transgenic Mice with Chosen-Site Integration. Proceedings of the
National Academy of Sciences - PNAS, 93(17), 9067–9072.
https://doi.org/10.1073/pnas.93.17.9067
Bunea, A. I., & Taboryski, R. (2020). Recent advances in microswimmers for biomedical
applications. Micromachines (Basel), 11(12), 1–24. https://doi.org/10.3390/mi11121048
Casanova, E. A. (2011). Molecular Mechanisms of Pluripotency in Murine Embryonic Stem
Cells. IntechOpen. https://doi.org/10.5772/14219
Das, A. T., Tenenbaum, L., & Berkhout, B. (2016). Tet-On Systems For Doxycycline-inducible
Gene Expression. Current Gene Therapy, 16(3), 156–167.
https://doi.org/10.2174/1566523216666160524144041
Efroni, S., Duttagupta, R., Cheng, J., Dehghani, H., Hoeppner, D. J., Dash, C., Bazett-Jones, D.
P., Le Grice, S., McKay, R. D. G., Buetow, K. H., Gingeras, T. R., Misteli, T., &
Meshorer, E. (2008). Global Transcription in Pluripotent Embryonic Stem Cells. Cell
Stem Cell, 2(5), 437–447. https://doi.org/10.1016/j.stem.2008.03.021
Gaj, T., Sirk, S. J., Shui, S. L., & Liu, J. (2016). Genome-editing technologies: Principles and
applications. Cold Spring Harbor Perspectives in Biology, 8(12).
https://doi.org/10.1101/cshperspect.a023754
Gumuskaya, G., Srivastava, P., Cooper, B. G., Lesser, H., Semegran, B., Garnier, S., & Levin, M.
(2024). Motile Living Biobots Self‐Construct from Adult Human Somatic Progenitor
Seed Cells. Advanced Science, 11(4). https://doi.org/10.1002/advs.202303575
Heisenberg, C.-P., & Bellaïche, Y. (2013). Forces in Tissue Morphogenesis and Patterning. Cell,
153(5), 948–962. https://doi.org/10.1016/j.cell.2013.05.008
Kattman, S. J., Witty, A. D., Gagliardi, M., Dubois, N. C., Niapour, M., Hotta, A., Ellis, J., &
Keller, G. (2011). Stage-Specific Optimization of Activin/Nodal and BMP Signaling
Promotes Cardiac Differentiation of Mouse and Human Pluripotent Stem Cell Lines. Cell
Stem Cell, 8(2), 228–240. https://doi.org/10.1016/j.stem.2010.12.008
59

Kim, Y.-G., Cha, J., & Chandrasegaran, S. (1996). Hybrid Restriction Enzymes: Zinc Finger
Fusions to Fok I Cleavage Domain. Proceedings of the National Academy of Sciences -
PNAS, 93(3), 1156–1160. https://doi.org/10.1073/pnas.93.3.1156
Lee, K. Y., Park, S. J., Matthews, D. G., Kim, S. L., Marquez, C. A., Zimmerman, J. F., Ardoña,
H. A. M., Kleber, A. G., Lauder, G. V., & Parker, K. K. (2022). An autonomously
swimming biohybrid fish designed with human cardiac biophysics. Science, 375(6581),
639–647. https://doi.org/10.1126/science.abh0474
Mo, C., Li, G., & Bian, X. (2023). Challenges and attempts to make intelligent microswimmers.
Frontiers in Physics, 11. https://doi.org/10.3389/fphy.2023.1279883
Morsut, L., & Gardiner, D.M. (2017). Regenerative Engineering and Developmental Biology:
Principles and Applications. CRC Press. https://doi.org/10.1201/9781315120188
Nawroth, J. C., Lee, H., Feinberg, A. W., Ripplinger, C. M., McCain, M. L., Grosberg, A.,
Dabiri, J. O., & Parker, K. K. (2012). A tissue-engineered jellyfish with biomimetic
propulsion. Nature Biotechnology, 30(8), 792–797. https://doi.org/10.1038/nbt.2269
Norrman, K., Fischer, Y., Bonnamy, B., Sand, F. W., Ravassard, P., & Semb, H. (2010).
Quantitative comparison of constitutive promoters in human ES cells. PloS One, 5(8).
https://doi.org/10.1371/journal.pone.0012413
Pagan‐Diaz, G. J., Zhang, X., Grant, L., Kim, Y., Aydin, O., Cvetkovic, C., Ko, E., Solomon, E.,
Hollis, J., Kong, H., Saif, T., Gazzola, M., & Bashir, R. (2018). Simulation and
Fabrication of Stronger, Larger, and Faster Walking Biohybrid Machines. Advanced
Functional Materials, 28(33). https://doi.org/10.1002/adfm.201804580
Papanayotou, C., & Collignon, J. (2014). Activin/Nodal signalling before implantation: setting
the stage for embryo patterning. Philosophical Transactions of the Royal Society of
London. Series B, Biological Sciences, 369(1657), 20130539–20130539.
https://doi.org/10.1098/rstb.2013.0539
Pauklin, S., & Vallier, L. (2015). Activin/nodal signalling in stem cells. Development
(Cambridge), 142(4), 607–619. https://doi.org/10.1242/dev.091769
Punovuori, K., Migueles, R. P., Malaguti, M., Blin, G., Macleod, K. G., Carragher, N. O., Pieters,
T., van Roy, F., Stemmler, M. P., & Lowell, S. (2019). N-cadherin stabilises neural
identity by dampening anti-neural signals. Development, 146(21).
https://doi.org/10.1242/dev.183269
Soncin, F., & Ward, C. M. (2011). The function of E-cadherin in stem cell pluripotency and
self-renewal. Genes, 2(1), 229–259. https://doi.org/10.3390/genes2010229
Swartz, M. A. (2003). Signaling in morphogenesis: transport cues in morphogenesis. Current
Opinion in Biotechnology, 14(5), 547–550. https://doi.org/10.1016/j.copbio.2003.09.003
60

Thomson, J. A., Itskovitz-Eldor, J., Shapiro, S. S., Waknitz, M. A., Swiergiel, J. J., Marshall, V.
S., & Jones, J. M. (1998). Embryonic Stem Cell Lines Derived from Human Blastocysts.
Science, 282(5391), 1145–1147. https://doi.org/10.1126/science.282.5391.1145
Touw, K., Hoggatt, A. M., Simon, G., & Herring, B. P. (2007). Hprt-targeted transgenes provide
new insights into smooth muscle-restricted promoter activity. American Journal of
Physiology: Cell Physiology, 292(3), C1024–C1032.
https://doi.org/10.1152/ajpcell.00445.2006
van den Brink, S. C., Baillie-Johnson, P., Balayo, T., Hadjantonakis, A. K., Nowotschin, S.,
Turner, D. A., & Arias, A. M. (2014). Symmetry breaking, germ layer specification and
axial organisation in aggregates of mouse embryonic stem cells. Development, 141(22),
4231–4242. https://doi.org/10.1242/dev.113001
van den Brink, S. C., & van Oudenaarden, A. (2021). 3D gastruloids: a novel frontier in stem
cell-based in vitro modeling of mammalian gastrulation. Trends in Cell Biology, 31(9),
747–759. https://doi.org/10.1016/j.tcb.2021.06.007
Wang, J., Greene, S. B., & Martin, J. F. (2011). BMP signaling in congenital heart disease: New
developments and future directions. Birth Defects Research. A Clinical and Molecular
Teratology, 91(6), 441–448. https://doi.org/10.1002/bdra.20785
Wang, R. N., Green, J., Wang, Z., Deng, Y., Qiao, M., Peabody, M., Zhang, Q., Ye, J., Yan, Z.,
Denduluri, S., Idowu, O., Li, M., Shen, C., Hu, A., Haydon, R. C., Kang, R., Mok, J.,
Lee, M. J., Luu, H. L., & Shi, L. L. (2014). Bone Morphogenetic Protein (BMP)
signaling in development and human diseases. Genes & Diseases, 1(1), 87–105.
https://doi.org/10.1016/j.gendis.2014.07.005
Wiedenheft, B., Sternberg, S. H., & Doudna, J. A. (2012). RNA-guided genetic silencing systems
in bacteria and archaea. Nature, 482(7385), 331–338.
https://doi.org/10.1038/nature10886
Woodgett, J., Batlle-Morera, L., Cohen, P., Doble, B., Wray, J., Nichols, J., Ying, Q.-L., & Smith,
A. (2008). The ground state of embryonic stem cell self-renewal. Nature, 453(7194),
519–523. https://doi.org/10.1038/nature06968
Wu, D., Yamada, K. M., & Wang, S. (2023). Tissue Morphogenesis Through Dynamic Cell and
Matrix Interactions. Annual Review of Cell and Developmental Biology, 39(1), 123–144.
https://doi.org/10.1146/annurev-cellbio-020223-031019
Yang, S., Hu, H., Kung, H., Zou, R., Dai, Y., Hu, Y., Wang, T., Lv, T., Yu, J., & Li, F. (2023).
Organoids: The current status and biomedical applications. MedComm (2020), 4(3).
https://doi.org/10.1002/mco2.274
Yan, W., Lin, C., Guo, Y., Chen, Y., Du, Y., Lau, W. B., Xia, Y., Zhang, F., Su, R., Gao, E.,
Wang, Y., Li, C., Liu, R., Ma, X., & Tao, L. (2020). N-Cadherin Overexpression
61

Mobilizes the Protective Effects of Mesenchymal Stromal Cells Against Ischemic Heart
Injury Through a β-Catenin–Dependent Manner. Circulation Research, 126(7), 857–874.
https://doi.org/10.1161/CIRCRESAHA.119.315806
Zallen, J. A., & Goldstein, B. (2017). Cellular mechanisms of morphogenesis. Seminars in Cell
& Developmental Biology, 67, 101–102. https://doi.org/10.1016/j.semcdb.2017.06.008
62

Abstract (if available)

Abstract Tissue engineering and biorobotics propel our understanding of cellular coordination, assembly, and interplay between form and function. Efforts have probed the capacity to use cellular components as actuators, sensors, and structural frames, integrating them with non-living material to create bio-hybrid machines. Of particular interest is the development of microswimmers, constructs that can swim at small scales with applications in drug delivery, bio-electronics, and ecological exploration. Recent efforts have explored the ability to engineer cellular-only constructs of microswimmers. The work aims to produce novel anatomic configurations and emergent functions from the intrinsic cellular machinery. Thus, pushing our knowledge of cell plasticity and collective behavior. Most recently, Gumuskaya et al., 2023 developed Anthrobots: self-organizing and cilia powered biobots of stem cells derived from adult human lung tissue.This thesis puts forth a new player in the realm of biobots and biological engineering: a self-motile cellular construct via cardiomyocyte actuation. Unlike previous constructs, this protocol entails the use of mouse embryonic stem cells that are coaxed towards a novel body plan and myocyte powered motility. The protocol is 10 days long, with emergence of contractility and motility by days 6 and 7 respectively, and consists of soluble factor delivery during specific temporal windows. The final constructs are 1019±101μm in length. In a motility assay on day 10 they display a displacement of 788.67±289.56μm. Further, I discuss developmental marker dynamics and morphological features that comprise the motile aggregates. Finally, I present an attempt to synthetically engineer aggregate body plans through cadherin mediated phase separation.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Engineering multicellular oscillators based on SynNotch-mediated intercellular dialogue

PDF

β-catenin couples self-renewal, induction and epithelial morphogenesis and patterning at the outset of mammalian nephrogenesis

PDF

Molecular role of EZH2 overexpression in Colorectal Cancer progression

PDF

Fibroblastic connective tissue cells: the blastema stem cells and source of large-scale chondrogenesis in the regenerating lizard tail

PDF

Nr5a2 balances lineage decisions to promote diverse connective tissue fates in the jaw

PDF

Characterization of metabolically distinct muscle resident fibro-adipogenic subpopulations reveals a potentially exploitable mechanism of skeletal muscle aging and disease

PDF

An image-based identification of aggressive circulating tumor cell subtypes

PDF

Dissecting the heterogeneity of mouse hematopoietic stem cells in vivo

PDF

Renal extracellular matrix bioactivity in a range of physiological states: novel insights for kidney regeneration

PDF

Investigating molecular roadblocks to enhance direct cellular reprogramming

PDF

Immunomodulatory and regenerative potential of amniotic fluid stem cells as a treatment strategy for pulmonary fibrosis

PDF

Skeletal cell fate plasticity in zebrafish bone development and regeneration

PDF

Understanding human nephrogenesis and scaling synthesis of organoids facilitate modeling of kidney development and disease

PDF

Identification of therapeutic targets in human cerebral brain organoid models of neurodegeneration

PDF

The role of ERK1/2 in mouse embryonic stem cell fate control

PDF

Regional localization and regulation of hematopoietic stem cells in the bone marrow stem cell niche

PDF

Encapuslated periodontal ligament and gingival mesenchymal stem cells in 3-D, injectable, and biodegradable scaffold: a unique platform for bone tissue engineering

PDF

Discovery of a novel role for Sonic Hedgehog during the early stages of large-scale murine rib regeneration

PDF

Pleotropic potential of Stat3 in determining self-renewal, apoptosis, and differentiation in mouse embryonic stem cells

PDF

Role of beta-catenin in mouse epiblast stem cell, embryonic stem cell self-renewal and differentiation

Asset Metadata

Creator Godage, Gaveen (author)

Core Title Derivation of cardiomyocyte-propelled motile cellular constructs via de novo developmental trajectory

Contributor Electronically uploaded by the author (provenance)

School Keck School of Medicine

Degree Master of Science

Degree Program Stem Cell Biology and Regenerative Medicine

Degree Conferral Date 2024-08

Publication Date 04/02/2025

Defense Date 06/25/2024

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

Tag biorobotics,developmental biology,OAI-PMH Harvest,tissue engineering

Format theses (aat)

Language English

Advisor Morsut, Leonardo (committee chair), Mariani, Francesca (committee member), Ying, Qilong (committee member)

Creator Email gaveen.godage@gmail.com,godage@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC11399BEDK

Unique identifier UC11399BEDK

Identifier etd-GodageGave-13569.pdf (filename)

Legacy Identifier etd-GodageGave-13569

Document Type Thesis

Format theses (aat)

Rights Godage, Gaveen

Internet Media Type application/pdf

Type texts

Source 20241002-usctheses-batch-1216 (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