Search

p300 and CBP are homologous proteins that possess histone acetyltransferase (HAT) activity. Both proteins play critical roles in various cellular processes including transcriptional activation. In this study, we aimed to investigate the effect of p300/CBP HAT inhibition on mouse embryonic stem cell (ESC) pluripotency and differentiation utilising the specific p300/CBP inhibitor, A-485. It was observed that p300/CBP catalytic inhibition reduced the growth of ESCs and elevated the proportion of cells in G1 phase of cell cycle. Transcriptome analysis following p300/CBP HAT inhibition revealed a significant alteration in gene expression at different time points. Transcriptional levels of pluripotency genes including Oct4, Nanog, Sox2, Esrrb among others were robustly downregulated. Interestingly, RNA-Seq analysis showed that several trophectoderm associated genes were upregulated in response to A-485 treatment. Moreover, these observations were accompanied by a reduction in acetylation levels of H3K18, H3K27 and H3K56 which are known substrates for p300/CBP. Assay for transposase-accessible chromatin using sequencing (ATAC-Seq) showed that chromatin accessibility was significantly changed following p300/CBP catalytic inhibition. Notably, the chromatin accessibility was reduced at the super-enhancer elements of pluripotency genes. To assess the effect of p300/CBP HAT inhibition on ESCs differentiation, a novel differentiation system known as gastruloids was utilised. Following p300/CBP inhibition, elongation of gastruloids was significantly blocked while expression of differentiation markers associated with elongated gastruloids was downregulated including Brachyury, Meox1 and Hox genes. Taking together, the data presented in this thesis suggests that p300/CBP HAT activity is indispensable for pluripotency and differentiation of ESCs.

Early mammalian embryos undergo remarkable changes in their metabolism, with a global transition from oxidative phosphorylation to glycolysis as they transit out of pluripotency and begin to make the first cell fate decisions in preparation for gastrulation. In humans this transition can be modelled with pluripotent cells in culture which exist in a naive (more pluripotent) and a primed (epiblast like) state and show the same differences in metabolic regulation. These two cell states also show major differences in their epigenetic landscape including in DNA methylation, repressive histone modifications, and spatial organisation of chromatin. I am interested in studying this metabolic regulation and its potential interface with epigenetic modifiers. In order to study metabolic regulation and heterogeneity within populations of naive and primed cells, I have adapted an existing method of constraint-based modelling to using constraints based on single cell RNA-seq data. The implementation of single cell models allows us to directly investigate the effect of transcriptional variability on the metabolism of the cells. Analysis of these modelling results shows that PCA of all reaction fluxes reveals clear differences between naive and primed cells, but interestingly also reveals subgroups within each cell type which exhibit different metabolic landscapes, which has not been previously observed. Notably, we have identified two genes which clearly identify these subgroups: SLC15A1 and SLC15A2. These genes encode transporters which are involved in uptake of dietary peptides, and there is no documentation of their function in pluripotency. Here we use several experimental techniques, including single molecule RNA fluorescence in situ hybridisation, to identify if these subpopulations present as our model suggests; future investigation may lead to novel insights into the regulation of these genes and their previously unnoticed role in stem cell metabolism and pluripotency.

The progressive degeneration of the A9 dopaminergic (DA) neurons of the substantia nigra lies at the heart of the pathology of Parkinson's disease (PD). Although current treatment options alleviate the symptoms, they do not prevent the cell death. As such, stem cell-based DA cell replacement therapies provide a promising treatment option, however little is understood about their immunogenicity. This PhD project investigates the immunogenicity of these cells and their implications for PD transplantation. My first aim was to understand the immunogenicity of embryonic stem cell (ESC)-derived midbrain DA progenitors (mDAps) in vitro. To address this, I studied the surface expression of several major immune molecules on these progenitors under normal and inflammatory conditions, given the latter would be the environment post-transplant. MHC-I expression was low in mDAps but was upregulated in inflammatory conditions; MHC-II and the co-stimulatory molecules always remained undetectable. To assess their immunogenicity in vitro, they were co-cultured with peripheral blood mononuclear cells (PBMCs) obtained from healthy blood, and mDAps evoked only little response from T cells. Such a response could though eventually lead to a graft rejection, which formed the basis of my second aim - to knockout (KO) MHC-I expression and to examine their differentiation ability and immunogenicity in vitro. B2M protein, that stabilizes the MHC class I complex, was knocked out using CRISPR-Cas9. These KO cell lines differentiated to authentic DA progenitors and neurons, and their expression of other immune molecules remained the same as their wild-type counterparts. The immunogenicity of these KO cells to T cells was also undetectable. Given that NK cells are toxic to cells lacking MHC-I, the response of NK cells was also assessed and KO mDAps were found to evoke no NK cell response in vitro. Finally, I looked at the effects of immunosuppressive drugs on mDAps survival and maturation in vitro given these agents will be given in the early phase clinical trials using these cells. I found that most of such agents did not affect the survival or maturation of the DA neurons, with the exception of mycophenolate mofetil. Overall, I found that human RC17 derived mDAps were no more immunogenic than human foetal ventral mesencephalic tissue and as such would not require any major change in strategy when used in clinical trials compared to those that have been done with primary human foetal tissue, although I further explored other aspects of this around looking at the effects of immunosuppressive agents on these cells as well as generated a MHC-I KO RC17 cell line.

Pluripotency is defined as the capacity to give rise to all cell types of the embryo proper. It arises in the early mammalian embryo but is lost after a short period of time as cells differentiate and become committed to different lineages. Prior to implantation, mouse epiblast cells enter the pluripotent naïve state, which can be captured in vitro in the form of embryonic stem cells. These cells are characterized by a capacity for indefinite self-renewal, and the ability to re-enter normal development upon being returned to the naïve epiblast. A complex transcription factor network promotes this state. Overexpression of one of many of these factors leads to stabilisation of the naïve state, with enhanced self-renewal and reduced spontaneous differentiation. However, the transcription factor Oct4 must be maintained within a tight window of expression; depletion results in extraembryonic differentiation, while overexpression also results in exit from pluripotency. Oct4 was identified as a protein expressed in the early embryo and in germ cells, and was subsequently discovered to be essential for the establishment of the naïve epiblast. In vitro studies determined that loss of Oct4 in ESCs induced trophoblast differentiation. Meanwhile, overexpression of Oct4 led to differentiation of ESCs, and constitutive expression of Oct4 was not sufficient to replace any of the extrinsic factors required for ESC self-renewal. Despite these dramatic phenotypes, the essential role of Oct4 remains unclear, further complicated by the finding that a reduced level of Oct4 promotes self-renewal at the expense of differentiation capacity. In this work, I generated a novel Oct4 fusion protein capable of rapid inducible degradation in order to study the immediate responses to removal of Oct4. This system utilizes the auxin responsive degradation domain of the Arabidopsis thaliana IAA17 protein to recruit a transgenic F-box protein Tir1 on addition of the small molecule auxin to the culture medium. Subsequent ubiquitination by the endogenous SCF complex leads to rapid proteolytic degradation of the Oct4 fusion protein, resulting in loss of detectable protein in as little as two hours. This system allows the study of immediate responses to loss of Oct4 in contrast to conventional depletion systems in which Oct4 levels decay over a protracted period making it difficult to disentangle direct and indirect effects. I established that several pluripotency-associated genes require Oct4 for their transcription. RNA levels of these factors decrease rapidly on depletion of Oct4, prior to significant changes in the expression of other key pluripotency factors such as Nanog and before protein levels of other factors can change dramatically. Furthermore, I established that the presence of Oct4 antagonises chromatin binding by a naïve transcription factor, revealing a possible mechanism by which increased Oct4 levels are detrimental to the naïve state. Together, these findings may be sufficient to explain the simultaneous requirement for, and antagonistic activity of, Oct4 in naïve pluripotent cells. I also found that an ESC level of Oct4 facilitates cell identity transitions. Cells constitutively expressing Oct4 and differentiated in vivo could be reverted to naïve pluripotency in vitro through the application of defined naïve pluripotency growth conditions. Additionally, in the course of these experiments I examined the phenotypic abnormalities that occur in mouse embryonic development under continuous expression of Oct4. In keeping with previous work, we observed abnormalities in limb development and in the skin. We also observed exencephaly in a number of embryos. In conventional exit from pluripotency, female cells must inactive an X chromosome in order to balance X-linked gene expression between males and females. This is achieved via expression of Xist from a single X chromosome, where it orchestrates chromosome-wide silencing. I show that Oct4 plays an important role in the regulation of Xist during the exit from pluripotency. Normally, Oct4 expression persists after the downregulation of most naïve transcription factors during early differentiation. I propose that this allows Oct4 to antagonise expression of the Xist, and thus ensure proper control over X chromosome inactivation. Together this work focuses on the dual roles of Oct4 in regulating both pluripotency and differentiation. I address the contradictory phenotypes relating to altered expression of Oct4, and establish a unifying theory to explain them. I put forward evidence that Oct4 promotes cell fate transitions by regulating naïve transcription factors. I propose that the environment plays a key role in determining cell identity, while Oct4 acts to maintain plasticity by preventing cells from being trapped within otherwise stable states.

Textbook accounts of vertebrate embryonic development have been based largely upon experiments on amphibian embryos, which have shown that the tissues of the trunk and tail are organised from distinct precursors that existed during gastrulation. In the mouse and chick, however, retrospective clonal analyses and transplantation experiments have demonstrated that the amniote body instead arises progressively from a population of axial precursors that are common to both the neural and mesodermal tissues of the trunk and tail. For this reason, they are known as neuro-mesodermal progenitors (NMps). Detailed studies of NMps have been precluded by their lack of a unique gene expression profile and the technical difficulties associated with isolating them from the embryo. Mouse embryonic stem cells (ESCs) provide the possibility of instead deriving them in vitro. ESCs have been used to model developmental processes, partly through large cellular aggregates known as embryoid bodies. These structures do not, however, resemble the axial organisation of the embryo and they develop in a disordered manner. This thesis presents a novel culture system of small, three-dimensional aggregates of ESCs (gastruloids) that can recreate the events of early post-implantation development, including axial elongation. Gastruloids are the first ESC-based model for axial elongation morphogenesis; this body of work characterises their development and identifies a candidate population of NMps within their elongating tissues. Additionally, this work establishes a xenotransplantation assay for testing the functional properties of in vitro-derived NMp populations in the chicken embryo and applies it to NMps from gastruloid cultures. The results of this assay show that gastruloids are a credible source of NMps in vitro and therefore offer a new experimental means to interrogate their properties. The use of gastruloids to recreate embryonic development has implications for basic research as a synthetic system and for the therapeutic derivation of other embryonic progenitors through bioengineering.

Embryonic stem (ES) cell maintenance and differentiation are dynamic processes controlled by various intrinsic and extrinsic factors. Identifying these factors will enhance the understanding about developmental process and improve the application of stem cells in clinic. Previous studies highlight a shift between non-oxidative and oxidative energy metabolism to play roles during differentiation. Oxidative metabolism is a major source of reactive oxygen species (ROS) which is regulated by a cytoprotective transcription factor, Nuclear factor erythroid 2-related factor 2 (Nrf2). Therefore, this study investigate relationship between metabolism, ROS, and Nrf2 during mouse ES cell differentiation. In vitro models representing early lineage differentiation were used. By measuring metabolic profiles, ROS, and Nrf2 levels from the models, Nrf2 was found related to pluripotency and ROS. However, relationship among metabolism and Nrf2 or ROS could not be detected. Gain- and loss-of-function experiments by pharmacological activator, short hairpin RNA knockdown, and CRISPR-Cas9 genome editing showed that Nrf2 could promote pluripotency and inhibit differentiation, especially during early differentiation toward neural lineage. This study suggested a new player in transcription control that governs pluripotency and differentiation.

The mRNA cap added to the 5’ end of nascent transcripts is required for the efficient gene expression in eukaryotes. In vertebrates, the guanosine cap is methylated at N7 position by RNMT, which is in complex with its activating subunit RAM. Additionally, the first and second transcribed nucleotides can be methylated at ribose O2 position by CMTR1 and CMTR2 respectively. The mRNA cap protects transcripts from degradation and recruits cap-binding factors to promote pre-mRNA processing, nuclear export and translation initiation. In mouse embryonic stem cells (mESCs), high levels of RAM maintain expression of pluripotency factors. Differentiation of mESCs to neural progenitors is accompanied by a suppression of RAM, resulting in downregulation of pluripotency factors and efficient formation of neural cells. Here, I demonstrated that the suppression of RAM during neural differentiation is promoted via ubiquitination and proteasomal degradation. Concurrently, neural differentiation is associated with an increase in CMTR1 expression, creating a developmental cap methyltransferase switch. Moreover, differentiation into endodermal and mesodermal lineages exhibited distinct changes in the mRNA capping enzymes expression. In mESCs, RAM promotes expression of translation-associated proteins and promotes global loading of mRNA on ribosomes. RAM contributes to the ESC-specific gene expression program, by maintaining optimal expression of pluripotency-associated transcripts and inhibiting expression of neural genes. Chromatin immunoprecipitation revealed that RAM, RNMT and CMTR1 promote binding of RNA polymerase II at gene loci. In RAM-repressed cells, RNA polymerase II binding was reduced at pluripotency-associated genes, but relatively increased at neural genes. Moreover, knock-down of RNMT or CMTR1 induced increased or decreased accumulation of RNA polymerase II at promoter proximal regions respectively. In naïve T cells, Rnmt or Cmtr1 conditional knock-outs caused downregulation of translation-related transcripts and upregulation of cell cycle transcripts. Furthermore, many transcripts were specifically dependent on RNMT or CMTR1 for expression, demonstrating distinct roles of these cap methyltransferases. Thus, the mRNA cap methylation emerges as an important regulator of pluripotency and differentiation, modulating gene expression at transcriptional and post-transcriptional levels.

A vast amount of gene expression data is available to biological researchers. As of October 2010, the GEO database has 45,777 chips of publicly available gene expression pro ling data from the Affymetrix (HGU133v2) GeneChip platform, representing 2.5 billion numerical measurements. Given this wealth of data, `meta-analysis' methods allowing inferences to be made from combinations of samples from different experiments are critically important. This thesis explores the application of localized pattern-mining approaches, as exemplified by biclustering, for large-scale gene expression analysis. Biclustering methods are particularly attractive for the analysis of large compendia of gene expression data as they allow the extraction of relationships that occur only across subsets of genes and samples. Standard correlation methods, however, assume a single correlation relationship between two genes occurs across all samples in the data. There are a number of existing biclustering methods, but as these did not prove suitable for large scale analysis, a novel method named `IslandCluster' was developed. This method provided a framework for investigating the results of different approaches to biclustering meta-analysis. The biclustering methods used in this work involve preprocessing of gene expression data into a unified scale in order to assess the significance of expression patterns. A novel discretisation approach is shown to identify distinct classes of genes' expression values more appropriately than approaches reported in the literature. A Gene Expression State Transformation (`GESTr') introduced as the first reported modelling of the biological state of expression on a unified scale and is shown to facilitate effective meta-analysis. Localised co-dependency analysis is introduced, a paradigm for identifying transcriptional relationships from gene expression data. Tools implementing this analysis were developed and used to analyse specificity of transcriptional relationships, to distinguish related subsets within a set of transcription factor (TF) targets and to tease apart combinatorial regulation of a set of targets by multiple TFs. The state of pluripotency, from which a mammalian cell has the potential to differentiate into any cell from any of the three adult germ layers, is maintained by forced expression of Nanog and may be induced from a non-pluripotent state by the expression of Oct4, Sox2, Klf4 and cMyc. Analysis of cMyc regulatory targets shed light on a recent proposition that cMyc induces an `embryonic stem cell like' transcriptional signature outside embryonic stem (ES) cells, revealing a cMyc-responsive subset of the signature and identifying ES cell expressed targets with evidence of broad cMyc-induction. Regulatory targets through which cMyc, Oct4, Sox2 and Nanog may maintain or induce pluripotency were identified, offering insight into transcriptional mechanisms involved in the control of pluripotency and demonstrating the utility of the novel analysis approaches presented in this work.

Embryonic stem (ES) cells are derived from the inner cell mass of the blastocyst stage embryo. They exhibit unlimited proliferation in culture and have the ability to differentiate into all three germ layers of the developing organism, a property defined as pluripotency. Previously it was reported that growth factor-bound protein 2 (Grb2) is required for differentiation of the epiblast, the embryonic tissue that harbours the pluripotent founder cells of the foetus. GRB2 is an adapter protein involved in the activation of the mitogen-activated protein kinase (MAPK) pathway in response to extracellular signals. It has also been implicated in the activation of the phosphoinositol-3-kinase (PI3K) pathway in response to fibroblast growth factor (FGF) signaling. The work presented in this thesis examines the role of Grb2 in ES cells and describes previously unreported contributions of this adaptor protein in regulating ES cell growth and differentiation. It has been previously been shown by others that Grb2 deficient (Grb2-/-) cells grow relatively normally in ES growth medium containing serum. However, in serum free conditions (N2B27 medium) in this project, proliferation of Grb2-/- cells is reduced compared with wild type and “restored” Grb2-/- cells stably expressing a Grb2 cDNA mini gene. Under serum free conditions, Grb2-/- cells grow in tight, refractive colonies. Nanog expression was uniformly upregulated, in contrast to the heterogeneous pattern reported in serum-based medium. Colony expansion on the substratum appears to be compromised, although there is no apparent defect in the initial attachment of Grb2-/- cells. Cell cycle analysis indicates that the slower growth of Grb2-/- cells in serum free medium could be due to lengthening of the G1 phase of the ES cell cycle. In an attempt to identify the signalling deficiency responsible for the growth defect of Grb2-/- cells, MAPK activation was restored by two methods, PMA a ligand that bypasses the requirement for Grb2, and Raf-ER, a conditionally regulated component of the MAPK pathway that acts downstream of Grb2 in the MAPK pathway. Although both approaches increased MAPK signalling they were unable to rescue the growth defect. This suggests that MAPK is not required or alone is not sufficient. Inhibition of Glycogen synthase kinase 3 β (GSK3 β ) is known to augment growth of ES cells under MAPK inhibition. Surprisingly, GSK3 β inhibition did not enhance Grb2-/- cell growth. Under GSK3 β inhibition, Grb2-/- ES cells fail to thrive. It is hypothesised that under these conditions cells undergo hyper-self-renewal at the cost of growth. Grb2-/- ES cells are reported to exhibit limited differentiation potential. To examine the potency of Grb2-/- cells, these cells were subjected to embryoid body (EB) and monolayer differentiation. Analysis of EBs showed a loss of Gata4, Gata6 and endoderm marker gene expression. However, markers of ectoderm (Sox1, Pax6, MAP2), the late epiblast/nascent mesoderm (Brachyury) and markers associated with gastrulation (Twist and Snail) were expressed. Outgrowths of morphologically and immunohistochemically identifiable neuronal cells confirmed differentiation of ectodermal cell types, indicating Grb2 is not required for neuronal differentiation. However, beating cardiomyocytes could not be identified in Grb2-/- EBs, though readily found in restored Grb2-/- cells expressing the Grb2 cDNA. This suggests that there is an essential role for Grb2 in the mesoderm/cardiomyocyte differentiation pathway. This may be due to a defect in GATA factor expression since these factors are essential for cardiogenesis. In serum-free monolayer differentiation, Grb2-/- cells formed neuronal cells. Additional inhibition of the MAPK pathway using a small chemical inhibitor failed to prevent this differentiation. However, biochemical analysis of the cells indicates that this occurs when ERK activation is very low, indicating differentiation was not MAPK-independent. Grb2 mediates FGF-MAPK induced exit from the naïve ground state. These data suggest a Grb2-independent pathway can also facilitate this transition. Grb2 is dispensable for differentiation in to some lineages. However as differentiation of Grb2-/- ES cells is restricted, this indicates Grb2 is required for true pluripotency.

Growth factor signaling is a key determinant of cellular decisions ranging from stem cell fate, to metabolic behavior. As such, gaining control over these signaling events is of critical importance to the field of regenerative medicine, which is continuously seeking novel means to tailor the cellular microenvironment to direct cells towards medically advantageous outcomes. Heparan sulfate (HS) glycosaminoglycans (GAGS) are sulfated polysaccharides found on the cell surface and in the extracellular matrix which are responsible for the engagement of growth factors as well as growth factor receptors as a means to spatiotemporally direct cell signaling events.Despite the obvious potential associated with controlling HS GAG-growth factor interactions, HS GAG-based approaches to manipulate cellular signaling has been minimal due to the complex nature of this class of sulfated polysaccharides. Sulfated GAGs like HS have a non-template driven synthesis, that is, their structure is not dictated by a genetic blue print. During assembly, the GAGs undergo a series of sulfations, isomerizations, and acetylations which give rise to their specific binding capacity, but it is this same complexity that poses a significant hurdle for synthetic or chemoenzymatic approaches geared at producing these structures for study and medical application. As a result, novel approaches must be developed to hone control over cell signaling using GAGs while circumventing the structural obstacle posed by the complexity of their structure. In this dissertation, I present a variety of accessible methods for the preparation of synthetic HS GAG mimetics, and demonstrate their efficacy in several contexts. In chapter 2, I introduce a HS GAG presenting, membrane incorporating, polymer which utilizes a multivalent display of commercially available HS GAG disaccharides to exert control over stem cell fate in a structure dependent manner. In chapter 3 I reveal the role of HS GAG in mesodermal differentiation. In chapter 4, I developed novel inhibitor of HS biosynthesis to arrest mESCs in a pluripotent state. Finally, in chapter 5, I reflect on the current state of HS mimetics and their use in stem cell biology.

Pluripotent stem cells reside transiently in the epiblast of blastocyst-stage embryos, until the proper time when they receive signals to differentiate into the germ layers that will give rise to the entire organism. However, the mechanisms that underlie the regulation of stem cell fate in the developing embryo, particularly how they maintain pluripotency, remain poorly understood. Here we show that embryonic stem cells (ESCs), the in vitro equivalents of stem cell populations within the epiblast, form and release extracellular vesicles (EVs), including microvesicles (MVs) and exosomes, to help maintain their stemness. Treating ESCs cultured under differentiation-inducing conditions with MVs and/or exosomes isolated from pluripotent ESCs blocked their differentiation and preserved stem cell characteristics including the ability to generate chimeric animals. Such effects were determined to be dependent on the extracellular matrix protein fibronectin associated with the vesicles, and its ability to bind to integrins and activate focal adhesion kinase (FAK) in recipient cells. These findings raise the exciting possibility that EVs might serve as a mechanism that helps stem cells within the blastocyst to maintain their pluripotent state.In addition to the roles that ESC-derived EVs potentially play in the physiological setting, we made interesting discoveries showing that they also can dramatically impact the behavior of differentiated cells by causing them to behave like stem cells, and even possibly induce pluripotency. The treatment of mouse embryonic fibroblasts (MEFs), a type of somatic cell, with MVs and exosomes produced by ESCs protected the cells from serum starvation-induced cell death and delayed senescence. Moreover, the EV-treated MEFs were also found to acquire the phenotypes of stem cells, such as the ability to form spheres, exhibit alkaline phosphatase activity, and express core stemness proteins Oct3/4 and Nanog. The EVs from ESCs strongly activated the cell survival and growth-promoting protein AKT in MEFs, and when EV-mediated AKT activation was inhibited, the MEFs no longer showed stem cell-like characteristics. Collectively, these findings shed light on how a unique form of intercellular communication can have important consequences in early development, as well as can potentially be used for regenerative medicine applications.

The forkhead box F1 transcription factor (FOXF1) is a mesenchymal-specific transcription factor and is expressed in mesenchyme-derived cells such as hepatic stellate cells and lung microvascular endothelial cells. FOXF1 functions are critical for transcriptional regulation during development of the liver, lung, and other organs, and has numerous roles in adult diseases. Here, we investigate FOXF1 and its role in hepatic fibrosis and its role in endothelial progenitor cells. Studying FOXF1 throughout different systems will enhance our knowledge base and can be used to develop further studies in a variety of systems. FOXF1 is expressed in the collagen-producing cells of the liver, the hepatic stellate cells (HSC), and HSC activation to myofibroblasts (MFs) requires FOXF1 presence. Therefore, our studies utilize a carbon tetrachloride-induced liver injury model to investigate the role of FOXF1 during liver fibrosis progression. We found that Foxf1 deletion increased collagen depositions and disrupted liver architecture. Timp2 expression was significantly increased in Foxf1-deficient mice while MMP9 activity was reduced. RNA sequencing of purified liver myofibroblasts demonstrated that FOXF1 inhibits expression of pro-fibrotic genes: Col1a2, Col5a2, and Mmp2 in fibrotic livers and binds to active repressors located in promotors and introns of these genes. Overexpression of FOXF1 inhibits Col1a2, Col5a2, and MMP2 in primary murine HSCs in vitro. In summary, we found that FOXF1 prevents aberrant extracellular matrix depositions during hepatic fibrosis by repressing pro-fibrotic gene transcription in HSCs and MFs. FOXF1 is additionally expressed in endothelial cells, which line blood vessels. FOXF1 in an important regulator of vascular formation and has roles in cellular proliferation. Therefore, our studies utilized endothelial colony forming cells (ECFC) to rescue mice with acute lung injury (ALI). We found a 53.8% increase in the survival of mice with ALI, which was associated with a decrease in lung injury; however, further evaluation of the injected ECFCs revealed no fluorescent tracer or FOXF1 expression. We therefore developed a novel embryonic stem cell (ESC) line with GFP:FOXF1 to both trace FOXF1 expression through differentiation to endothelial progenitor cells (EPC) in vitro and to trace the cell in vivo for future studies. Our novel differentiation method yields ~95.2% endothelial cells, 51.5% of which are FOXF1-positive. In summary, we have successfully developed a novel ESC line and a novel EPC differentiation protocol to be used in future studies.

The processes of cell proliferation, differentiation, migration, and self-organization during early embryonic development are governed by spatially and temporally varying morphogen signals. Such dynamic morphogen patterns drive cell fate specification at the proper location and time. However, current in vitro methods typically do not allow for precise, dynamic, spatiotemporal control of morphogen signaling and are thus insufficient to readily study how morphogen dynamics impact cell behavior. In embryonic stem cell (ESC) models for early development, for example, analogous tissue patterns spontaneously emerge but mechanistic insight into this self-organization has been limited. In this dissertation, we combine engineered methods for optogenetic stimulation and biological analysis to photoactivate Wnt/B-catenin signaling and to study morphogen dynamics and self-organization of human pluripotent stem cells. First, we show that optogenetic Wnt/B-catenin pathway activation can be controlled at user-defined intensities, temporal sequences, and spatial patterns in human ESCs (hESCs). We expressed a fusion of the plant blue-light photoreceptor Cryptochrome 2 (Cry2) to the Wnt co-receptor LRP6 (the optoWnt system) in hESCs and induce Cry2-LRP6c oligomerization and subsequent canonical Wnt/B-catenin activation using novel engineered illumination devices for optogenetic photostimulation and light activation at variable amplitudes (LAVA). Such LAVA devices are a LED-based, programmable illumination system for photostimulation of multi-well plates that can be readily incorporated into the workflow of routine cell culture and allow controlled and quantitative spatiotemporal light patterning. We optimized the LAVA board optical configuration for illumination uniformity and achieve programmable photostimulation of independent wells of 24-well or 96-well culture plates kept in standard 37C tissue culture incubators. Each well can be wirelessly programmed through a graphical user interface (GUI) at user-defined intensities (0 -- 20 uWmm-2, 0.005 uWmm-2 resolution), temporal sequences (10 ms resolution), and spatial patterns (100 um resolution). We demonstrate LAVA board performance by modulating the intensity, timing, and spatial location of canonical Wnt/$\beta$-catenin signaling in hESC cultures using the optoWnt optogenetic system. We show that Wnt pathway activation and hESC differentiation is dose-responsive to light intensity and duration of illumination, and that spatial patterning can be used to simulate the embryonic, spatially polarized presentation of the Wnt ligand. In parallel, we engineered a microscope capable of dynamic photostimulation and high-resolution imaging with structured illumination microscopy (SIM). A digital micromirror device (DMD) is used to project user-defined photostimulation sequences onto the sample with diffraction-limited spatial resolution (800 nm). For fluorescence imaging, a series of multi-spot patterns is projected onto the sample to reconstruct a high-resolution SIM image with a 2-fold resolution improvement over the diffraction limit. Such multi-color, user-defined photostimulation and SIM imaging (opto-SIM) allows live-cell timelapse imaging while localizing Wnt/B-catenin photoactivation to specific regions within hESC cultures. Unlike LED-based illumination devices, such as the LAVA board system, the opto-SIM system enables pattern projection at subcellular resolution. Furthermore, projected patterns can be dynamically updated based on feedback from measurements acquired during experiment progression. The opto-SIM system also enables temporal control of signaling at a resolution of 60 Hz, as well as 8-bit greyscale modulation to achieve dose-dependent optoWnt activation. Such flexibility in spatiotemporal patterning gives the unprecedented ability to stimulate stem-cell based model systems, organoids, or embryos with time-varying patterns of morphogen gradients. Lastly, we implement optogenetic control of canonical Wnt signaling to determine whether differential Wnt signaling can model human gastrulation and lead to emergence of organized shape and structure through collective cell rearrangement. We achieve optogenetic control of Wnt signaling in hESCs by illuminating hESC cultures expressing the optoWnt system with LAVA devices. Wnt/B-catenin signaling was activated over a high dynamic range (>500-fold) and drove broad transcriptional changes and mesendoderm differentiation of human ESCs at high efficiency (>95% cells). Furthermore, activating Wnt signaling in subpopulations of ESCs in 2D and 3D cultures induced cell self-organization and morphogenesis reminiscent of human gastrulation, including changes in cell migration and epithelial to mesenchymal transition. We thus developed an hESC model for studying Wnt-mediated morphogenesis in early development. Using this gastrulation model in combination with transcriptomic analysis and single-cell migration studies, we show that Wnt signaling is sufficient for inducing self-organization of cells in an EMT-dependent manner. Our findings thus reveal an instructive role for Wnt in directing cell patterning in this ESC model for early embryogenesis. In summary, we developed and utilized engineered methods for spatial and temporal light patterning to photoactivate Wnt signaling in a precise location and at a given time within embryonic stem cell cultures. Such optogenetic activation of morphogen signaling in specific cell subpopulations allows studies of cell-cell interactions and signaling dynamics that regulate early embryonic development.

The ability to culture stem cells in vitro has provided access to a wide variety of cellular states that are inaccessible or too rare to study otherwise. Here we demonstrate the use of stem cells to model two distinct developmental transitions, and examine the role that epigenetic regulation of transcription plays in controlling stem cell fate and function.First, we examine the transition from mouse embryonic stem cells (mESCs), representative of the pre-implantation epiblast, to mouse epiblast stem cells (mEpiSCs), representative of the post-implantation epiblast. These two states, while maintained by distinct signaling pathways, both retain the ability to re-integrate into their respective tissue of origin and contribute to the development of the entire organism. While the transcriptome of the two cell types is largely similar, they rely on dramatically different sets of enhancers to regulate that expression. 97% of the genes that are shared between the two states undergo a change in enhancer usage in the transition. The enhancers that arise in the mEpiSC state are present, but inactive, in the preceding mESC state and become enhancer clusters downstream in development.Second, we examine the development of oligodendrocytes from oligodendrocyte progenitors (OPCs). Oligodendrocytes wrap neural axons in a lipid-rich sheath known as myelin. This myelin is required for proper signal conduction, but is subject to immune attack in multiple sclerosis (MS). OPCs replace damaged oligodendrocytes early in the course of disease, but eventually fail. A chemical genetics screen reveals that BET bromodomain proteins are required for oligodendrocyte development. These proteins are epigenetic readers that integrate chromatin state and other signals to allow transcription elongation at many genes. Other inhibitors of elongation show similar effects on development, and treatment with BET bromodomain inhibitor (S)-JQ1 blocks activation of oligodendrocyte genes. Elongation genes are enriched at MS risk alleles, and these alleles show evidence of enhancer function. In MS tissue, elongation genes are disrupted, and Hexim1 is increased in oligodendrocyte lineage cells.In combination, our results demonstrate the broad role that the epigenetic regulation of transcription plays in modulating cellular identity and function in stem cells, both in development and disease.

Hematopoietic stem cells (HSCs) are able to self-renew, generate all lineages of the blood, reconstitute the hematopoietic system upon bone marrow (BM) transplantation, and thereby cure diseases of the blood and immune system. BM transplantations have successfully treated a wide range of blood-related diseases (i.e., β-thalassemia, leukemia, and HIV/AIDS). However, current demand for BM transplantations outstrips the quantity of HSCs available through cord blood banks and BM registries, and further requires finding matching HLA-types between patients and donors. Given these restrictions in quantity and compatibility, there is significant interest in developing alternative sources of transplantable HSCs by differentiating HSCs in vitro from pluripotent stem cells (PSCs), such as embryonic stem cells (ESCs). Unfortunately, lack of understanding of the regulatory mechanisms governing HSC function has prevented their in vitro generation. The focus of my thesis research is on deciphering the molecular blocks preventing the generation of HSCs from ESCs, and identifying key molecular cues to help overcome these blocks. In the course of my research, I identified the medial HOXA genes as critical regulators for human HSC self-renewal, and explored the pathways that regulate the HOXA genes that are dysregulated during ESC differentiation cultures. Our lab developed a two-step differentiation protocol mimicking the environments a developing HSC would encounter in the embryo and successfully generated cells with the human HSC immunophenotype, CD34+CD38-CD45+CD90+GPI80+. However, although ESC-derived cells differentiated into definitive erythroid, myeloid, and T-cells, they displayed impaired self-renewal and engraftment ability compared to hematopoietic stem/progenitor cells (HSPCs) isolated from the fetal liver (FL), the site of the most active HSC expansion and differentiation during human development. Microarray analysis witnessed the overall close molecular correlation between ESC-HSPCs and FL-HSPCs, but revealed the lack of expression of the medial HOXA genes in ESC-derived cells as a critical defect preventing their self-renewal. Knockdown of HOXA5 or HOXA7 in FL-HSPCs recapitulated the self-renewal defects observed in ESC-HSPCs. A six-day stimulation of the retinoic acid (RA) signaling pathway was sufficient to significantly upregulate HOXA gene expression in ESC-HSPCs at day 6. However, HOXA gene expression was not sustained long-term.Microarray and ChIP data further revealed that the regulator of HOXA genes, MLL1, was expressed, but not bound to HOXA genes in ESC-derived cells, suggesting the inability to recruit MLL1 to HOXA genes prevents their activation during ESC differentiation. I identified in the human genome a region of 89% sequence homology to the mouse lincRNA Mistral, which in mice recruits MLL1 to HOXA6/HOXA7 genes. RNA-seq and microarray data showed high expression of putative human MISTRAL in FL-HSCs but not in their differentiated progeny or in ESC-derived hematopoietic cells, and high correlation of HOXA6/A7 expression to Mistral expression in multiple cell types. These data identify insufficiency of HOXA gene expression is a developmental barrier for generating HSCs from pluripotent cells and suggest RA-signaling is a major inductive signal and MISTRAL a novel component of the fetal HSC regulatory machinery that conveys “stemness” in hematopoietic cells.

The rate of glycolytic metabolism changes during differentiation of human embryonic stem cells (hESCs) and reprogramming of somatic cells to pluripotent stem cells. However, the functional contribution of glycolytic metabolism to pluripotency is unclear. Here we show that the degree of pluripotency is associated with glycolytic rate, whereby naive hESCs exhibit increased glycolytic flux, MYC transcriptional activity, and nuclear localization of N-MYC relative to primed hESCs. This is consistent with the inner cell mass of human blastocysts which exhibit increased MYC transcriptional activity relative primed hESCs and elevated nuclear N-MYC levels. Reduction of glycolysis decreases self-renewal of naive hESCs and feeder-free cultured primed hESCs, but not primed hESCs grown in feeder-supported conditions. Reduction of glycolysis in feeder-free primed hESCs also enhances neural specification. These findings reveal associations between glycolytic metabolism and the state of pluripotency, differences in the metabolism of feeder- versus feeder-free cultured hESCs, and identify methods for regulating self-renewal and initial cell fate specification of hESCs.

Transforming Growth Factor – beta signaling plays important, pleiotropic roles in embryonic development as well as tissue homeostasis. TGF-beta signaling dose is tightly regulated in a context-specific manner and even subtle perturbations result in a spectrum of disease phenotypes of varying severity. In the canonical TGF-beta pathway, TGF-beta signals are mediated by activated receptor-regulated Smads (R-Smads, Smad2/3) that form a complex with Smad4 and participate in transcriptional regulation of TGF-beta-mediated genes with sequence specific DNA binding factors and context-specific co-regulators. Both Smad2/3 and Smad4 regulate signal duration and intensity differentially. This is consistent with the variability and spectrum of phenotypes observed upon their conditional deletions in mice, with Smad2/3-associated phenotypes being more severe than those associated with Smad4. However, precise mechanisms that modulate the functional and phenotypic differences attributed to Smad2/3 and Smad4 are poorly understood. Tripartite motif containing (Trim) 33 - a co-regulator of both activated Smad2/3 and Smad4, is a candidate gene for understanding fine tuning mechanisms of TGF-beta signaling. Here, I have investigated the role of Trim33 in early embryogenesis. My results show that Trim33 controls visceral endoderm differentiation in vitro and pre-cardiac mesoderm differentiation in vivo. Epiblast-specific Trim33 mutants die during the second half of gestation as a result of ensuing cardiac failure. My findings imply that Trim33 is required for appropriate early mesendoderm differentiation soon after its induction, regulates TGF-beta superfamily signaling in a biphasic manner in the pre-cardiac mesoderm, and that in more committed cell lineages Trim33 is dispensable.

Glycosaminoglycan (GAG) mediated growth factor signaling initiates differentiation in embryonic stem cells (ESCs). Manipulation of the GAG interactome affords a novel tool to understand and direct stem cell differentiation. In this study, a high-throughput miniaturized cellular microarray platform, consisting of gelatin robotically spotted onto passivated glass slides is established to assess the effects of cellular glycan microenviroment on growth factor signaling in mouse embryonic stem cells (mESCs). Heparan sulfate (HS) acts as co-receptor for fibroblast growth factor 2 (FGF2) and the subsequent activation of the mitogen activated protein kinase (MAPK) pathway is measured by extracellular regulated kinase (ERK) phosphorylation. Using HS deficient EXT1-/-¬ ESCs, FGF2 signaling is rescued using soluble and immobilized heparin. Immobilization of glycans on cellular microarrays offer potential as a powerful tool for systematically evaluating and directing cellular differentiation.Then, using conventional culture conditions, the interactions of HS and FGF2 are screened using bis-2-methyl-4-amino-quinolyl-6-carbamide, or surfen, and compared to the commercial FGFR inhibitor PD173074. Surfen reversibly antagonizes heparan sulfate in the µm range and can HS mediated interactions can be rescued using soluble heparin, a highly sulfated HS analog.

Apoptosis is a highly conserved form of programmed cell death in multicellular organisms where a cell activates caspase proteases to trigger its own demise. It has critical functions under physiological conditions in removing superfluous or damaged cells from a healthy organism. However, dysregulated apoptosis plays an important role in diseases: improper or elevated apoptosis leads to diseases of cell loss, such as diabetes or neurodegeneration, while deficiencies in apoptosis can contribute to tumor cell growth and survival.While the importance of apoptosis during late embryogenesis is well-established, its function during the earliest stages of development, namely the exit of embryonic stem cells (ESC) from the pluripotent state, is unclear. Here, I discovered that apoptosis plays a critical role in removing slow-differentiating murine ESC from the total cell population. This is initiated by p53-dependent upregulation of the Fas death receptor on straggling ESCs, which then triggers apoptosis specifically in these cells. An inability to initiate apoptosis, by transient knockdown or genetic deletion of components of this pathway, causes retention of slow-differentiating ESCs and a global delay in differentiation, both in vitro and in an in vivo teratoma model. Thus, apoptosis is crucial in promoting the efficient differentiation of ESCs.While apoptosis is important for normal development, elevated levels due to cellular stress can cause inappropriate cell loss and degeneration, leading to disease. One such example is retinitis pigmentosa, an inherited disease where rod photoreceptors are progressively lost from the neural retina, eventually leading to blindness. There is increasing evidence that accumulation of unfolded proteins within the endoplasmic reticulum (ER) of rod photoreceptors causes ER stress and subsequent cell death. Here, I utilize a novel small molecule inhibitor to show that inhibition of IRE1α, an ER transmembrane kinase/endoribonuclease that can signal both adaptation and apoptosis in response to ER stress, preserves rod photoreceptor viability and function in two rodent models of ER stress-induced photoreceptor degeneration.These findings reveal new insights into the role of apoptosis under both normal and stressed conditions, and may have implications for human diseases such as cancer and neurodegeneration.

The Piwi proteins are a clade of the Argonaute family specific to animals that bind small RNAs called Piwi-interacting RNAs (piRNAs). The Piwi/piRNA system is indispensable in germ cells, where it has dual roles in germline stem cell self-renewal and the epigenetic silencing of retrotransposons. Although Piwi/piRNA activity is generally assumed to be restricted to the germline, there is mounting evidence that Piwi genes and proteins also have a somatic function. In primitive metazoans, Piwi orthologues appear to be important for self-renewal of somatic stem cells, and Piwi homologues are expressed in multiple mammalian tissues as well as in a variety of cancers. I hypothesised that the Piwi/piRNA system is important for mammalian somatic stem cell self-renewal, in particular, in maintaining pluripotency in embryonic stem cells. I also hypothesised that the Piwi/piRNA pathway is involved in conferring the stem cell-like phenotype of certain cancer cells. To test these hypotheses, I investigated the expression and function of the murine Piwi homologue, Miwi2, in embryonic stem (ES) cells and a panel of breast cancer cell lines. In this thesis, I show that Miwi2 is expressed in the cytoplasm of murine ES cells, and that its expression diminishes along with ES cell differentiation; however, Miwi2 does not appear to be critical for ES cell self-renewal and pluripotency. I also show that Miwi2 is highly expressed in metastatic 4T1 murine breast cancer cells, and that ectopic Miwi2 expression in syngenic, non-metastatic 4TO7 cells causes gene and small RNA expression changes in pathways relevant to metastasis. In ES cells, I find that Miwi2 associates with repeat-derived piRNAs and piRNAs produced from large intergenic clusters, as well from the 3’UTR of genes, including the master regulators of pluripotency, Oct4, Sox2 and Nanog. In 4T1 cells, Miwi2 associates with piRNAs from 3’UTR clusters, as well with 5’ tRNA fragments that are linked to translational control. The function of Miwi2 in ES cells and cancer biology remains unclear, but my data indicate that Miwi2 and its associated piRNAs have a role in the regulation of gene expression in somatic cells.

Our ancestors and our children are linked by a single, special cell lineage called the germline. Germ cells are tasked with the role to accurately pass DNA from one generation to the next. In today's society infertility is an important health concern as it is estimated to affect 10% of the reproductive age population. In many cases infertility can be traced to abnormal germline cell development. The early events of germline formation are difficult to study because of limited materials, especially low germ cell numbers in the process of embryogenesis.To overcome the challenge of limited materials, an in vitro method to derive cell types faithfully reporting in vivo phenotypes needs to be devised to facilitate the research process. The key results from the in vitro model should then be evaluated carefully in vivo to show relevance in development. My thesis first focused on establishing an in vitro model to recapitulate early development of germline, by differentiating embryonic stem cells (ESCs) to form in vitro primordial germ cells (iPGCs), in a dish. This method allows good number of germ cells to be produced for molecular and biochemical studies. Using the in vitro model, key germline modifiers for ESC maintenance and germ cell formation were identified, one of which is protein arginine methyltransferase 5 (PRMT5). PRMT5 is a type II arginine methyltransferase (PRMT) that modifies symmetrical dimethylated arginines (SDMAs) on proteins, which substrates include histone H2A and H4, as well as the Sm proteins involved in RNA splicing. However, the molecular function of PRMT5 in ground state ESCs remains unknown. In our study, for the first time, we generated an inducible knock out of Prmt5 in ESCs cultured in ground state pluripotency (cultured with 2i inhibitors and leukemia inhibitory factor). We characterized PRMT5's role in ESCs through the method of paired-end RNA sequencing to show that PRMT5 affects cell survival, proliferation and chromatin organization etc., by affecting RNA spicing. By generating a conditional knock out to specifically deplete PRMT5 in the germline in vivo, we showed that PRMT5 is necessary for mammalian PGC formation. Taken together, we developed a differentiation system to ask questions about key factors regulating ESC and germline development. One of the key factors we identified is PRMT5. The role of PRMT5 was scrutinized using paired-end RNA sequencing in the inducible knock out ESCs, suggesting that PRMT5 is necessary to ensure normal splicing in "RNA processing", "DNA damage response" and "chromatin modification" for cell to survive. Finally, we utilized in vivo mouse model to validate events that happen in vitro, showing that PRMT5 is indeed a survival factor both for ground state pluripotency and mammalian germline.

Pluripotency establishment and maintenance are critical for both the development of mammalian embryos and the production of pluripotent stem cells. Investigations into the molecular mechanisms underlying pluripotency have revealed multiple signaling networks that generate diverse pluripotent states. The relationship between these networks, their upstream transcription factors, and their associated pluripotent states is complex and poorly understood in both pluripotent stem cells and embryos. In this thesis I examine three aspects of pluripotency regulation: the role of maternal transcription factors in embryogenesis, novel effects of genetic background and naivety on differentiation to placental stem cells, and the contribution of known stem cell factors to regulation of pluripotency in vivo.

The basis of developmental biology lies in the idea of when and how cells decide to divide or to differentiate. Previous studies have established several signaling pathways that determine cell fate decisions, including Notch, Wingless, Hedgehog, Bone morphogenetic protein, and Fibroblast growth factor. Signaling converges on transcriptional factors that regulate gene expression. In mouse embryonic stem cells, I explored how pluripotency and differentiation are regulated through opposing actions of beta-catenin-mediated canonical Wnt signaling, and the mechanisms underlying Sonic hedgehog signaling in generating progenitor cells in the ventral neural tube.

Unlimited self renewal capacity and differentiation potential make human pluripotent stem cells (PSC) a promising source for the ex vivo manufacture of red blood cells (RBC) for safe transfusion. Current methods of erythropoiesis from PSC suffer from low yields of RBCs, most of which contain embryonic and fetal rather than adult hemoglobin. We have previously shown that homo-dimerization of the intracellular component of MPL (ic-MPL) induces erythropoiesis from human cord blood progenitors through unique, EPO-independent mechanisms. The goal of the present study was to investigate the potential of ic-MPL dimerization to induce erythropoiesis from human embryonic stem cells (hESC) and to identify the signaling pathways activated by this strategy. I present here evidence that ic-MPL dimerization induces EPO-independent erythroid differentiation from hESC through AKT signaling, by generating erythroid progenitors and promoting definitive erythropoiesis with increased RBC enucleation as well as increased gamma:epsilon globin ratio and production of beta-globin protein. ic-MPL dimerization is significantly more potent than EPO in inducing erythropoiesis from hESC and activates a network of key erythroid factors involved in apoptosis, cell cycle, and differentiation. These findings open up potentially new approaches toward the generation of therapeutically relevant RBCs for transfusion.

Heart failure is a leading cause of death in United States. One of the causes of heart failure is associated with the death or loss of cardiomyocytes (CMs). Since adult CMs do not regenerate, their death permanently compromises myocardial contractile function. Stem cell transplantation is one therapeutic strategy to replace damaged or lost myocardial tissue to restore cardiac function. Embryonic stem cells (ESCs) are an attractive population for cardiac repair because they can self-renew unlimitedly and differentiate into all cell types including CMs. Furthermore, ESC derived CMs can functionally integrate with the recipient organ and improve heart function after transplantation. However, a major challenge in ESC-based cardiac therapies is that the differentiation efficiency of ESCs into CMs has been very low (~1%). And large numbers of cells are required for administration for each patient. In this study, a protocol for efficient generation of CMs from hESCs was explored by optimizing various staged components in the microenvironment. Specifically, I 1) developed Honeycomb Microwell chips to generate homogeneous EB for CM differentiation; 2) optimized Actvin A/BMP4 concentration for CM differentiation; 3) optimized effects of extracellular matrix (ECM) signaling and investigated some mechanisms of ECM signaling on CM differentiation. The optimized protocols reproducibly generate approximately 70% CMs from H7 and H9 hESCs. These hESC derived CMs can now be enriched and tested for their ability to enhance cardiac function in preclinical animal models and for utility in drug discovery for future study.

The spatial organization of the genome is linked to its biological function. However, the relationship between specific gene regulatory networks that govern cell identity and large-scale organization of genomes remains unclear. To investigate the basis of distal chromatin interactions occurring between genomic regions mega-bases away on the same or different chromosomes, we mapped long-range chromatin interactions in embryonic stem cells (ESCs) using 4C-seq, and examined the genomic features of the interacting regions. Confirming previous results, we show that open, accessible versus closed chromatin character is the primary determinant of distal chromatin interaction preferences, where interacting regions exhibit very similar open/closed chromatin character. We extend these results by demonstrating that genomic regions highly enriched for binding by the pluripotency transcription factors Oct4, Sox2, and Nanog preferentially co-localize, as do regions strongly enriched for Polycomb proteins and trimethylation of histone H3 at lysine 27 (H3K27me3), which include the Hox clusters. Consistent with a spatial segregation of these transcriptional networks, we find that Nanog and Polycomb proteins occupy distinct spaces in the nucleus. Importantly, loss of the Polycomb protein Eed and H3K27me3 diminishes the preferential interactions between regions normally highly enriched for Polycomb proteins and H3K27me3 without dramatically changing long-range chromatin interactions related to the open/closed chromatin state. Finally, a comparison of interactomes in ESCs and fibroblasts uncovered an ESC-specific spatial organization to the mouse genome that is gradually re-established upon reprogramming to induced pluripotent stem cells (iPSCs). Together, our data suggest that transcriptional networks that govern ESC identity play a role in determining genome-organization. We propose the existence of a hierarchy in the organization of chromatin contacts wherein, at the largest scale, the open/closed chromatin character defines an interaction space and overall chromosome conformation; on a finer scale, cell type-specific transcriptional networks direct preferential distal interactions, which we speculate are critical for efficient regulation of transcription and establishment of local chromatin environments.

Sperm and oocytes are terminally differentiated, sex-specific germ cells, which, upon fertilization will generate a new embryo and leads to species propagation by sexual reproduction. Though fated only to generate eggs or sperm, germ cells have the unique property to imbue zygotes with totipotent capacity, which facilitates the formation of all tissues the embryo will need to survive to adulthood. These characteristics are facilitated by germ cells' ability to pass on genetic information to the next generation, as well as their capacity to initiate genome-wide reorganization and removal of epigenetic information inherited by germ cells during embryogenesis. It is hypothesized that remodeling of this epigenetic information is essential to drive proper embryo development. While these events are not well understood, it is known that the events that underlie these unique properties are initiated in early development, shortly after the germ line is established as a pool of primordial germ cells (PGCs). Efforts to unravel these mechanisms that underlie totipotency in germ cells have been limited due to the inability to isolate, study, and manipulate PGCs. To overcome this obstacle, we hypothesized that PGCs cells could be differentiated from pluripotent embryonic stem cells, and that these cells would serve as a surrogate cell type for the study of PGC biology.Establishing a new model of lineage differentiation from embryonic stem cells required the development of assays and criteria to rigorously test identity, developmental staging, and epigenetic progression to determine if an in vitro model is able to recapitulate features of endogenous PGCs. To accomplish this, we developed a scalable and transgene-free method to differentiate immature PGCs in vitro using the cell surface markers SSEA1 and cKit that are developmentally and epigenetically reminiscent of immature PGCs. We applied existing assays to validate PGC identity, and devised a new stringent assay based on genetic deletion of a known PGC determinant. We developed a single-cell gene expression methodology to compare gene expression signatures of in vitro derived PGCs and endogenous PGCs, and identified novel criteria to define PGC identity from early endogenous PGCs and in vitro-generated PGCs.We next used in vitro PGC differentiation to investigate genome-wide DNA demethylation, one of the first epigenetic reprogramming events undertaken by early PGCs. By combining the scalability of this differentiation system with next generation methylation sequencing techniques, we generated the first DNA methylation maps of in vitro derived PGCs, and determined with sequence-specific information that DNA demethylation is genome-wide and likely to involve loss of DNA methylation as a consequence of cell division. We also investigated potentiators of active DNA methylation loss, including the Tet proteins, and their roles in early PGC development.Finally, we applied single cell gene expression technology to define developmental progression of human PGCs isolated from the gonads of fetuses from elective terminations. We identified a common progenitor stage of PGC development in the human fetal gonad. Furthermore, we adapted our single cell gene expression approaches to interrogate differentiation strategies in the generation of the common human PGC progenitor in vitro. Together, we have developed a differentiation system to ask questions about epigenetic progression in early germ cells, and have utilized single cell gene expression technology and genomics to characterize seminal events in the epigenetic reprogramming of human and mouse germ cells.

The ultimate goal of regenerative medicine is to create functional engineered biomaterial to repair or replace the damaged tissues. My work firstly was to perform multi-scale characterization on the collagen hydrogels used as scaffolds. Different initial collagen concentrations and incubation temperatures were identified to affect the properties of collagen hydrogels, such as the microstructure obtained with second harmonic generation imaging (SHG), nanostructure imaged with transmission electron microscopy (TEM) and molecular structures detected with Raman spectroscopy. Comparing to the real tissues, the drawback of collagen hydrogels is the lack of mechanical strength. Cross-linking has been known to increase the mechanical strength. I then characterized the effect of different cross-linking reagents including glycation reagents, genipin and 1-Ethyl-3 (3dimethylaminopropyl) carbodiimide (EDC) on the collagen hydrogel. SHG imaging showed that glycation reagents and genipin modify the microstructure within collagen hydrogel but EDC has no effect on the microstructure. TEM images revealed that glycation reagents and genipin also remove the native-like striation pattern in nanostructure but EDC has no effect on it. Raman spectroscopy showed that glycation reagents and EDC affect the molecular information of collagen hydrogels. Lastly, two different types of stem cell, mouse embryonic stem cell (ES cell) and induced pluripotent stem cells (iPS cells) were seeded to the collagen hydrogels and then examined the effect of different structure of collagen hydrogel on the neuronal differentiation of stem cells. The behaviors of G-Olig2 ES cell and a7wt iPS cell were compared during early differentiation with retinoic acid into a neural lineage employing separately the encapsulated or topographic 3D collagen hydrogel models. For the encapsulated model, the differentiation of ES cell and iPS cell were slower for the cells encapsulated within 4 g/l collagen hydrogels compared to the cells in 2 g/l hydrogels. The differentiation process of both cells types seeded on top of the 3D collagen hydrogels was affected by the structure modified by cross-linking. For example, the differentiation rate of ES cells was slower on the cross-linked hydrogels but the differentiation rate of iPS cells was faster on the cross-linked hydrogels.

In eukaryotic cells, a key regulatory step in target gene expression is the recognition of the core promoter by the TFIID complex. TFIID is composed of the TATA-box binding protein (TBP) and TBP-Associated Factors (TAFs) (Goodrich and Tjian, 2010; Hochheimer and Tjian, 2003; Naar et al., 2001). To understand gene regulation at the core promoter, I have been focusing on one component of the core promoter recognition complex-TAF3 that was originally identified as a subunit of the TFIID complex in HeLa cells (Gangloff et al., 2001). It was later found that, while other TFIID subunits are destroyed during myogenesis, TAF3 is selectively retained in myotubes in a specialized complex with TBP-related factor 3, TRF3 (Deato and Tjian, 2007) and the sub-nuclear localization of TAF3 serves as another potential mechanism to influence transcription during myogenesis (Yao et al., 2011). Intriguingly, TAF3 recognizes trimethylated histone H3 lysine 4 (H3K4me3) (Vermeulen et al., 2007), which is associated with actively transcribed genes and also with silent developmental genes that are poised for activation upon ES cell differentiation (Mikkelsen et al., 2007). In our recent studies, surprisingly it was found that TAF3 is highly enriched in ES cells. More importantly, during in vitro EB formation, which mimics early embryogenesis, TAF3 protein levels were selectively reduced. However, the levels of other canonical TAFs and TBP remained stable. Functional and gene expression analysis revealed that high levels of TAF3 are dispensable for ES cell self-renewal but are required for endoderm lineage differentiation and to prevent premature specification of neuroectoderm and mesoderm. Genome-wide binding studies by ChIP-seq found that unlike other canonical TFIID subunits such as TBP and TAF1 that bind only promoters, TAF3 targets both promoters and distal enhancer-like sites in ES cells.Importantly, TAF3 sites were further clustered into four distinct classes based on the context of co-localization with other factors. Briefly, as well as TAF3, class 1 regions are promoter regions enriched for TFIID, Pol II and H3K4me3. Their proximity (< 1kb) to the TSS of known genes confirms that they are predominantly TFIID-bound promoters. By contrast, Class 2 regions have low levels of TFIID and H3K4me3 but are enriched for Oct4/Nanog/Sox2 and mediator components. Class 3 regions are specifically enriched for TAF3, CTCF and cohesin subunits. Finally, class 4 regions are not enriched for any of the factors we considered besides TAF3. Regions from the last three Classes are generally far away (~10's kb) from the TSS.Further studies revealed that genes down-regulated upon TAF3 depletion are selectively associated with regions enriched for TAF3, CTCF and cohesin (Class 3) and TAF3 only (Class 4), while no such association was seen for genes up-regulated upon TAF3 depletion. Thus, it's likely that class 3 and class 4 regions are required for the efficient expression of genes in their vicinity and depletion of TAF3 interferes with this function. Indeed, our parallel biochemical experiments confirmed a direct protein-protein interaction between that TAF3 and CTCF. This novel interaction was further shown to mediate DNA looping between promoter distal sites and core promoters to regulate proper transcription activation. Notably, we also found that the class 2 binding regions are enriched around up-regulated genes upon TAF3 depletion, consistent with a mechanism of transcription repression by TAF3 in association with Oct4/Nanog/Sox2. Together, our data support the model that TAF3 orchestrates a complex long-distance chromatin interaction network that safeguards the finely-balanced transcriptional programs underlying ES pluripotency.

Here we present a microfluidic culture platform for enhancing single stem cell survival. Traditional plate culture is inadequate for large scale single cell studies because of (1) less than 40% single stem cell survival leading to (2) possible inaccuracies in the biostatistics in single cell studies from lack of sufficient data. This platform mitigates these issues by doubling overall survival rates as well as increasing the available data points by two orders of magnitude, both factors which improve statistical certainty. The platform is fabricated using widely accepted biocompatible polymers and incorporates the novel integration of an enclosed microwell array with an upstream linear gradient generator. The device can selectively trap single cells, tightly control intercell spacing, and change the chemical microenvironment around the stem cells in real-time. In this dissertation, we first characterize the performance of the platform in improving single cell survivability. Subsequently, the system was tested using two difference stem cell types: mouse embryonic stem cells (mES) and induced pluripotent stem cells (iPS). Using this platform we are able to highlight the thresholds of leukemia inhibitory factor concentrations which lead to metastable gene expression of Nanog, a key stem cell pluripotency regulator in mES cells. The enhanced survival of single cells has also enabled the statistically significant observation of Nanog+, non-proliferative, single mES cells which are previously unreported in literature.

When an embryonic stem cell (ESC) differentiates, it must both silence the ESC self-renewal program as well as activate new tissue-specific programs. In the absence of DGCR8, a protein required for microRNA (miRNA) biogenesis, mouse ESCs are unable to silence the ESC self-renewal program during differentiation. Screening by reintroduction of all known miRNAs one at a time into Dgcr8 -/- ESCs in differentiation-inducing conditions enabled the identification of numerous miRNAs which silence the ESC self-renewal program. Expression levels of many of these miRNAs are induced during ESC differentiation. Of these miRNAs, most are expressed in specific cell types whereas a single family, the let-7 family, is broadly expressed across differentiated cell types. In various assays of ESC self-renewal, let-7 family miRNAs rescue the inability of Dgcr8 -/- ESCs to silence self-renewal. However, let-7 miRNAs failed to silence self-renewal in wild-type ESCs, suggesting that ESC-expressed miRNAs inhibit the capacity of let-7 to silence self-renewal. Indeed, introduction of the embryonic stem cell cycle regulating (ESCC) miRNAs blocked the capacity of let-7 to induce silencing of self-renewal in Dgcr8 -/- ESCs. mRNA profiling and bioinformatic analysis showed that let-7 and ESCC miRNAs function in part through opposite regulation of Myc transcription factors and Lin28. The opposing regulation of these factors contributes to a network, which reinforces the switch from a self-renewing to a differentiated cell state. These results suggested that additional screen positive miRNAs function in similar antagonistic networks with ESCC miRNAs. Indeed, introduction of the ESCC miRNAs prevented the additional screen positive miRNAs from silencing self-renewal in Dgcr8 -/- ESCs. mRNA profiling and bioinformatic analyses suggest that screen positive miRNAs and the ESCC miRNAs oppositely regulate multiple molecular pathways including the G1/S cell cycle transition. Inhibition of the G1/S transition in wild-type ESCs promotes loss of markers of ESC self-renewal. These findings suggest that miRNAs through destabilization of the ESC cell-cycle may promote loss of ESC self-renewal during differentiation. These studies show that different classes of miRNAs positioned in the context of complex biological networks function to either promote or antagonize ESC self-renewal.

The mitochondrion is an important and essential cell organelle which provides about 90% energy for the organism and plays a number of important roles in cell functions. Recent research found that mitochondrial distribution in somatic cells is homogenously scattered throughout the cytoplasm, while the distribution in some stem cells and mature oocytes is perinuclear. These findings suggested that the spatial distribution of mitochondria may be related to their normal functions. In addition, Bavister has hypothesized that this periuclear localization of mitochondria may indicate the pluripotency of stem cells. But mitochondrial distribution has not been examined in embryonic stem cells to date. My thesis investigates the distribution of mitochondria during development in mammalian cells and proposes experiments to determine and compare the spatial distribution of mitochondria in embryonic stem cells and differentiated cells using fluorescence staining of MitoTracker Green to test Bavister’s hypothesis. In addition, research suggested that mitochondrial distribution is mediated by the cytoskeleton in higher eukaryotes. Therefore, an additional goal of my research was to address the effects of cytoskeletal disruptors on mitochondrial arrangement.

The oocyte (egg) is the mother of all cells, propagating genetic and epigenetic information, integrating paternal components, and reprogramming the fertilized embryo to generate offspring. However, the oocyte is mysteriously short-lived. Female reproductive potential is limited in mammalian species including humans where cessation of fertility is reached by mid-age, and 10-15% of couples are infertile due to female factors in half of these instances. Because they are restricted in number, differentiation of oocytes from embryonic stem cells (ESCs) will facilitate analyses of the genetic, epigenetic, and environmental factors affecting oocyte development and contributing to infertility and/or birth defects. The differentiation of functional oocytes from autologous embryonic stem cells will ultimately enable the understanding, potential treatment, and/or prevention of infertility in women. Additionally, ESC-derived oocytes may be used as nuclear transfer recipients to study cellular reprogramming and early embryo development.ESCs can differentiate into cell types of three somatic germ layers and the germline. Although putative germ cells with oocyte-like characteristics were previously reported to spontaneously differentiate from mouse ESCs in vitro, functional analyses and correlation to endogenous oocyte development in vivo have been limited. In this study, I developed germ cell-specific reporter and surface marker strategies to isolate and characterize mouse ESC-derived germ cells. Using these strategies, I established a developmental timeline and genetic program of female germ cell differentiation from ESCs in vitro that initially paralleled endogenous oocyte development in vivo. However, ESC-derived oocyte maturation eventually failed, as complete meiotic progression and ovarian follicle formation were not detected. To overcome this in vitro maturation bottleneck, I examined endogenous mouse fetal oocyte development following transplantation to develop a synchronized ovarian niche that was competent to direct oocyte maturation. I then transplanted ESC-derived oocytes into this ovarian niche and analyzed transplants for oocyte maturation in ovarian follicles. Indeed, ESC-derived oocytes functionally integrated into the ovarian niche, recruited endogenous somatic granulosa cells, and successfully developed to the primary follicle stage. In summary, this study provides rigorous evidence of physiologically relevant ESC-derived oocyte identity and function, and presents a potential clinical strategy for treatment of infertility through ovarian niche-based transplantation.

Gene modification of hematopoietic stem cells (HSCs) has the potential to cure genetic, malignant and acquired diseases. Despite success in pre-clinical gene therapy studies, achieving genetic correction or a therapeutic response in humans has been challenging. HIV-1-based lentivirus vectors have come to the forefront of pre-clinical studies due to their ability to more effectively transduce quiescent HSCs. Drug resistance gene expression coupled to chemotherapy after HSC transplantation may support in vivo selection of gene-modified cells while protecting the patient from chemotoxicity. We hypothesized that lentivirus-mediated transfer of a methotrexate (MTX) resistance gene, Tyr22-dihydrofolate reductase (Tyr22-DHFR), into stem cells would support long-term stable gene expression in vivo and protect hematopoietic daughter cells from MTX toxicity. To test our hypothesis, we first generated high-titer lentivirus vectors expressing Tyr22-DHFR and green fluorescent protein (GFP) in different genetic configurations, and then compared MTX resistance, enzyme activity and GFP fluorescence in mouse and human cell lines including human embryonic stem cells (hESCs). Tyr22-DHFR-HSCs protected transplanted mice from MTX myelotoxicity, and conferred a significant survival advantage compared to MTX treated GFP-HSC transplanted mice. To assess the feasibility of a physiologic scale-up in a large animal model, we demonstrated DHFR-GFP expression in canine CD34+ cells and long-term engraftment of gene-modified cells in vivo. MTX administration increased gene-marking in the peripheral blood of one dog, without causing cytopenia. We also defined the optimal priming of HSCs (c-G-CSF/c-SCF BM), transduction conditions and MTX tolerated doses in dogs. Finally, we present a novel application of selective expansion of hESCs-derived cells in mouse xenografts. Methotrexate-resistant (MTXr)-DHFR hESCs gave rise to MTXr-GFP+ teratomas, indicating that that gene-modified cells retain their pluripotency during MTX treatment. MTXr-hESCs placed in stromal cell co-culture differentiated into GFP+ hemato-endothelial cells, including CD34+CD45+ subsets, which subsequently gave rise to MTXr-hematopoietic colony forming cells (CFCs). Finally, we showed that MTX administration of mice bearing hESC xenografts supported in vivo selection of Tyr22-DHFR-hESC-hematopoietic cells and increased engraftment of gene-modified cells in the bone marrow of treated mice. Taken together, these results show that lentivirus vectors effectively transduce MTXr-DHFR into HSCs, thereby preventing life-threatening myelotoxicity (as observed in our mouse studies), and supporting long-term engraftment of gene-modified cells in vivo. These studies mark significant progress of MTX resistance gene therapy toward clinical trials in humans.

Embryonic stem (ES) cells possess unlimited self-renewal ability and the potential of differentiating to all tissue organs, making ES cells the ideal source for cell therapy and drug screening. However, present culture techniques are both expensive and labor intensive, unable to meet the projected high demand for ES cells. Three-dimensional culture is capable of mimicking in vivo growth environment and therefore, cells show their native morphology and functions. 3-D culture is superior to 2-D culture by eliminating the necessity of gelatin coating and frequent subculture. The STO cells conditioned medium is capable of replacing expensive cytokine leukemia inhibitory factor (LIF). Dynamic culture of ES cells in spinner flasks showed better cell growth and maintenance of pluripotency. Further scale up to a two stage perfusion culture system was successful, providing higher cell expansion and Oct-4 and SSEA-1 expressions. Human ES cells expansion in perfusion bioreactor also reached the success. hES cells grows faster in FBB than on 2-D surface. The hES cells pluripotency was also maintained well. However, neural differentiation efficiency was lower in 3-D culture than in 2-D system due to the higher possibility of causing de-differentiation in 3-D system. Astrocyte conditioned medium was superior to retinoic acid in inducing neural differentiation. Hydrodynamics and its influence on cell attachment and growth in 3-D cultures were also studied. Increased mixing intensity enhanced both cell attachment and detachment rates. Cell detachment kinetics can be described by Bell model. The interaction between Kolmogorov eddy size and the scaffold’s pore size played a vital role in affecting shear damage to cells growing in the scaffold. When Kolmogorov eddy size is smaller than the pore size, shear damage to cells is apparent. Mathematical modeling of the fibrous bed bioreactor (FBB) was investigated too. Axial dispersion model can represent the mass transfer in ordered disc packed FBB. However, both radial and axial dispersion plus convection contribute significantly to the mass transfer in spiral wound FBB. Packing methods showed different performance in different scales. Cell aggregate size should be controlled less than at 170 µm if there is 30% DO.