Biophysical Determinants of Early Embryonic Stem Cell Fate Specification

As an embryo progresses through early stages of development, the position of each cell is precisely defined so that it is provided with the appropriate cues to differentiate into the required cell type. We are just beginning to learn the nature of many of these cues, and it is clear that they include both signals secreted and taken up by cells and also signals sensed by cells as a result of how they are tethered to each other and to their surrounding tissues. While classic cellular responses to cues involve a specific receptor on the cell’s surface responding to a specific extracellular cue, the signals that cells sense as a result of their adhesion properties are more complex. We are working to understand how cells sense and respond to the forces imparted on them by other cells and by their underlying substrate, and how this affects their ability to differentiate into cell types of interest. We have developed a method to grow human embryonic stem cells on substrates of different stiffness, and have found that this dramatically changes how they organize within colonies and how they then respond to developmental cues. Specifically, when we provide cues to embryonic stem cells to initiate differentiation towards the mesoderm lineage, we find that cells on soft substrates respond much more robustly and differentiate more efficiently than cells on stiff substrates. Mesoderm forms during the developmental process known as gastrulation, and we also see aspects of this complex process recapitulated in our system, specifically when cells are grown in embryo-sized colonies on soft substrates. For example, we observe cells migrate and ingress into a region with similarities to the gastrulation-initiating structure called the primitive streak. Based on these results, we are designing methods and tools to more comprehensively understand how forces are set up within the differentiating colonies and correlate regions of high or low cell stresses with expression of proteins that are crucial for setting up the cells for efficient differentiation. Using a technique called monolayer stress microscopy, we can monitor how tension fields develop over time in primed and differentiating colonies. We have evidence that cell-cell contacts are strengthened when cells are grown on soft substrates, and that these are required to allow the cells to efficiently differentiate. We have also found regional expression of differentiation markers in the primitive-streak-like regions. We therefore intend to combine these observations to understand how colony organization affects cell stresses, and how these affect subcellular protein localization, which ultimately affects how cells respond to soluble extracellular cues to determine their developmental fate. This work will ultimately provide us with a better understanding of the fundamental processes by which cells respond to extracellular cues, including how soluble signals synergize with mechanical and tissue-level signals, allowing us to apply these principles to design optimized differentiation systems that mimic the endogenous environments in which cells differentiate during embryogenesis.

Human embryonic stem cells (hESc) or their induced pluripotent (iPSc) counterparts have the potential to be differentiated into all cells of the body and thus hold out the promise of replacement cell therapies for medical applications. Understanding the “how to” of the multiple stages of this process is critical for the success of these goals, which must be optimized both for the identity and yield of the target cell. In our CIRM Basic Biology Grant we have been investigating how the softness or stiffness of the surface on which the hESc are adhered and growing impacts the efficiency of the first step in the differentiation process, in which hESc assume one of three intermediate identities referred to as endoderm, mesoderm, or ectoderm. Specifically we are studying the transition to mesoderm the intermediate that subsequently gives rise to heart and skeletal muscle, bone, blood, and connective tissues among other medically useful types. We received our grant on the strength of the preliminary observation that surfaces with a softness similar in magnitude to structures of the developing embryo predispose hESc to make this transition with dramatically higher efficiency than when grown on much harder surfaces akin to those used routinely for hESc culture. Beginning in Year 1 and culminating now in Year 2 we have now identified the basis for this increase in efficiency to be published shortly in a prestigious scientific journal: Cell Stem Cell. In brief β1 integrins, which form part of the protein connection between the hESc and the surface are activated more on harder surfaces and ultimately lead to destabilization of β catenin, a key protein component of the protein connection called an adherens junction (AJ) that joins one cell to another. β-catenin also serves as a key molecular driver of mesoderm commitment when relocated to the cell nucleus where it turns on genes involved in this commitment. We show that the source of β-catenin is largely that which is present in AJ and is released when AJ breakdown shortly after receiving experimental signals to differentiate to mesoderm. Hard surfaces therefore impede this accumulation, impede AJ formation, and deprive the cell of sufficient β-catenin to provide an unambiguous mesoderm differentiation signal. The opposite occurs when hESc are cultured on a surface with the softness of the embryo. We show that a variety of different molecules acting both before and after receiving mesoderm inducing signals are involved in the pathway that stabilizes β-catenin on soft surfaces, including the components of the AJ itself and amplified local production of wnts, proteins released from the cell that prevent AJ released β-catenin from destruction. In other words soft surfaces as opposed to hard surfaces set up conditions for self –reinforcing feedforward amplification of the signal that exceeds a threshold for mesoderm commitment. This has practical implications for understanding this first commitment step and how it may be improved, provides insight into the mechanisms that underlie a widely used but poorly understood method of differentiation using hESc; namely releasing hESc from surface attachment, and also provides a framework for understanding how mesoderm might form in the embryo. In Year 2 we have also advanced related work centered on the observation that this mesoderm differentiation is not random in collections of hESc on soft surfaces but occurs in specific regions of hESc collectives near their free edges. As with many other cells types hESc possess protein machinery that in addition to allowing one cell to stick to another or to a surface also allow them to pull against these anchor points. In conjunction with collaborators in Spain we have shown that those regions that differentiate into mesoderm are also associated with the highest pulling forces. In Year 3 we are continuing efforts to try to understand if there is a causal connection between these two events. For example, such forces may promote the assembly or disassembly of AJ and thereby influence the strength of mesoderm commitment signals. Our ongoing studies are aimed at mapping in finer spatiotemporal detail the association between these mechanically active regions and mesoderm commitment. In Year 2 we have strong evidence that we may have succeeded in developing a difficult but important tool that we anticipate will allow us to accomplish this goal. Specifically, we have inserted a reporter into the hESc genome that can be observed in living cells downstream of, and linked to one of the earliest mesoderm commitment genes. This will allow us to identify cells in real time early in the commitment process and map their positions and movements relative to the underlying mechanical forces.