A. Stem cell systems biology and bioengineering
Mouse (m) embryonic stem cells (ESC), extracted from the inner cell mass of a pre-implantation blastocyst, were first isolated via co-culture on mouse embryo derived feeder layers in 1981. Seven years later it was demonstrated that robust feeder-free cultures could be maintained in the presence of serum and the cytokine leukemia inhibitory factor (LIF). LIF signalling is propagated by causing the phosphorylation of the transcription factor signal transducer and activator of transcription (STAT) 3. Over the last five years we have developed mathematical models of signalling in mESC and demonstrated, for the first time, that signalling thresholds, auto-regulatory feed forward control loops, and the organization of cells into colonies are determining factors in the rate and trajectories of pluripotent cell differentiation. This work is significant because it demonstrates that cellular organization, signalling and cell fate are linked parameters. We have used this understanding to develop new clinically- and commercially-relevant culture technologies for the controlled manipulation of mouse (and human) pluripotent cell fate. More recently, it was discovered that mESC self renewal can be maintained in a defined serum-free medium by the addition of bone morphogenetic protein (BMP)4. BMP4, a transforming growth factor (TGF)-β family member, signals through the Smad family of transcription factors, specifically Smad1 . Human (h) ESCs were first isolated by James Thomson in 1998. However, unlike mESCs, hESCs do not appear to require LIF/STAT3 or BMP4/Smad1 for self-renewal. They instead, as we and others have shown, self-renew in the presence of Activin A, another TGF-β super-family member, and basic fibroblast growth factor (bFGF) which activate Smad2 and extracellular signal-regulated kinase (ERK), respectively. This difference in signalling to maintain an apparently similar core transcriptional network consisting primarily of Oct4, Sox2, and Nanog raises the intriguing question of how two different signalling pathways converge on similar targets in two similar apparently pluripotent cell types. Species (mouse vs. human) differences have been commonly implicated to dismiss this incongruousness. However, recently an Activin A (TGF-β) / bFGF-responsive Oct4, Sox2, and Nanog positive mouse cell has been isolated from a post-implantation blastocyst. These cells have been termed “EpiSCs” due to their similarities to ESCs but also to the fact that they are derived from and appear similar to later stage epiblast cells (the cells directly responsible for giving rise to the three germ layers ectoderm, mesoderm and endoderm). In this project, we continue our studies into how signalling network dynamics and local cellular microenvironments are integrated to control cell fate. We are specifically interested in quantitatively examining the transition between ESC and EpiSC, and understanding the underlying signalling mechanisms that stabilize the pluripotent cell state. We are developing new strategies, based on the reprogramming work of Yamanaka, to quantitatively examine the ESC to EpiSC transition in both directions. Our goal is to understand how exogenous signalling integrates with the core pluripotency network and how this network can be reactivated in post-pluripotent cells.

The clinical utility of hematopoietic (blood) stem cells (HSCs) arises from their ability to engraft and sustain multilineage hematopoiesis in compromised hosts. Umbilical cord blood (UCB) is a very attractive source of HSCs. The major limitation of the use of UCB for hematopoietic transplantation is the limited number of stem and progenitor cells obtained in each UBC collection. In this research program we are exploring the hypothesis that HSC output during in vitro culture is governed by intercellular (between cell) protein signalling networks and that culture manipulations such as media dilution and cell subpopulation selection will perturb this self-organizing network, and can be systematically applied to improve culture performance. We have previously demonstrated that mature (lin+) cells produced in culture secrete negative regulators of HSC expansion (TGF-β1 and MIP-1α), and the secretion of these regulators can be modulated independently via media dilution and cell-population removal. We have developed automated systems to manipulate subpopulations of cells and to control the delivery of liable proteins to culture. In all of this work the predominant observation is that blood stem cell growth in vitro [and perhaps also in vivo] is limited by feedback networks from differentiated cells. Our next step in this project is to investigate the hypotheses that a) the blood forming system can be depicted as a cell-cell interaction network, linked through nodes representing soluble factors, b) that the dynamics of blood stem cell differentiation (rapid development of mature blood cells at the expense of blood stem or progenitor cell recovery) provides negative feedback that limits stem cell self-renewal and c) that specific receptor inhibitors used to target control points in cell-cell interaction networks that regulate blood stem cell fate. To accomplish this we will combine experimental and computation approaches to understand how specific blood stem cell niche components (different cell types and matrix proteins) interact in isolation and as integrated unit. Systems-biology research to date has primarily focused on elucidating the topological features and dynamics of intra-cellular networks, with the implicit assumption of cell populations as homogenous, autonomous units. Inter-cellular communication networks, represented with cell types as vertices, and functional (rather than molecular) interactions as edges, have been largely unexplored. We aim to reconstruct, in silico, cell-cell interaction networks, and examine the behavior and properties of such networks. Our approach will be to incorporate increasingly available databases on chemical binding and receptor expression based signals into cell-cell interaction networks. The ability to move from descriptive network depictions to constraint-based modeling approaches has been very successfully used in metabolic engineering to drive the design of new “synthetic” organisms to produce bio-products. We propose to use similar approaches, for the first time in regenerative medicine, to help us understand the key controlling molecules and cells in blood stem cell integration and endogenous regeneration. Finally, we will take advantage of our established human UCB automated bioreactor system to test model predictions and, ultimately new ways to manipulate blood stem cell responses. This project should help to define the role of the blood cell-cell “interactome” in blood stem cell fate control and could underpin mechanistically based technologies to grow blood stem cells ex vivo or to target them in vivo fate using mAbs and small molecules.

Pluripotent (P) and embryonic (E) stem cells (SC) are of significant interest as a source of therapeutically useful cells. ESC differentiation in stirred or high-density cultures has been problematic, thus preventing clinically or industrially relevant cell production. Several years ago we developed novel ways to culture mouse ESC aggregates at high cell densities in homogeneous and controllable bioreactor systems by microencapsulating them in agarose gel microdroplets. When placed in stirred suspension bioreactors, encapsulated mESC could be used to produce scalable quantities of blood, endothelial, and cardiac cells in a controlled environment. An advantage of this system is that it allows for the investigation of physiochemical parameters (oxygen, glucose) on ESC development, as well as the further engineering of ESC differentiation. Our focus over the next few years will be to develop novel technologies to control human pluripotent stem cell expansion and differentiation in integrated bioreactor systems. A design criterion in these studies is to make hESC propagation and differentiation robust, developmentally relevant and amenable to scalable cell production. To this end, we are exploring methods to use novel biomimetic strategies (gradients and scaffold-mediated signalling) to influence hESC differentiation trajectories, and combining microfabrication technologies and bioreactors to generate of hESC aggregates and cells in the quantities necessary for biotechnological applications. Recent funding from the Juvenile Diabetes Research Foundation (JDRF) has focused our efforts on the development of an integrated expansion and differentiation system for pancreatic endoderm production. Type 1 diabetes results from the autoimmune destruction of the insulin-producing pancreatic islet b-cells, and produces uncontrolled blood sugar levels and a host of associated complications including blindness, renal failure, and cardiovascular disease. Although remarkable progress has been made in the transplantation of cells derived from cadavers, the need for cells far outweighs the quantity available. One solution to this problem is to develop novel technologies to generate functional pancreatic cells, or their progenitors, in vitro via the controlled differentiation of pluripotent stem cells such as human embryonic stem cells. To be clinically useful such a technology should be capable of routinely producing large number (as many as 5x108 to 5x109 functional cells). It is with this goal in mind that we have formed a new interdisciplinary collaboration with the Stem Cell Biology lab of Gordon Keller. Recently, several studies, including those from the Keller laboratory, have demonstrated success in producing of insulin-expressing cells and/or their progenitors from pluripotent cells. While these studies confirm that the controlled differentiation of hESCs along the pancreatic lineage is possible, the goal of producing clinically relevant numbers of transplantable cells has not yet been reached. Overcoming this limitation will require both a mechanistic understanding of how culture conditions impact the proliferation of undifferentiated stem cells, and how cytokines can be efficiently and effectively delivered to induce, at high yields, the differentiation of pluripotent cells along the pancreatic lineage. In this project will develop novel bioprocess engineering strategies to expand human pluripotent stem cells (based on aggregate size control, identify the major culture related impediments in the efficient generation of pancreatic endoderm from human pluripotent cells, and utilize microwell-based screening strategies to identify cell-cell interactions and molecules useful for the maturation of pancreatic endoderm into functional beta cells. This research will not only define approaches that dramatically increase the availability of pancreatic cells for use in transplantation and drug discovery, but will also inform our understanding of the cellular and molecular mechanisms by which pluripotent cell commit to specific lineages.