Quantitative time-lapse analysis of pluripotency and trophectoderm differentiation at the single-cell level
Embryonic Stem Cells (ESCs) have the ability to change, or differentiate, into a diverse set of specific cell types (placental cells, neurons, muscle, etc.). Alternatively, they can also grow and divide to make more ESCs (self-renew). These abilities are the basis for their potential usefulness in medical applications. However, a population of ESCs usually does not differentiate into a single cell type. When apparently identical cells are placed together in the same culture dish, they typically generate a mixture of different cell types. This variability can arise from small variations in the local environment of each cell, from signals cells use to communicate with one another, or from intrinsic randomness (“stochasticity”) of the biochemical reactions that cells use to control differentiation. In order to manipulate ESCs it is crucial to understand the origins of this heterogeneity. Recent work has made it clear that cellular decisions result not from the control of a single gene, but rather from the interaction of many genes and proteins. In the case of ESC development, some of the genes and interactions that control differentiation decisions are now known. However, it is unclear why they act differently in different cells. Traditional techniques average over large cell populations, making it difficult to address this issue. Because of the intrinsic variability in ESC differentiation it is crucial to study it in individual cells. To this end, we will develop a microscopy-based system to simultaneously follow multiple genes in individual cells. We will create ESCs that contain three differently colored fluorescent protein labels that can be used to track the dynamics of important genes. We will record ‘movies’ of individual cells and their levels of gene expression as they grow and change into other cell types. Using this data in conjunction with mathematical models we will determine how the behavior of the system as a whole varies from cell to cell, how it responds to particular perturbations, and how it can be manipulated. More specifically, we will analyze the genetic system that allows ESCs to avoid differentiation (that is, to remain in the ESC state), as well as the system that allows them to change into placental cells. In this effort, we will directly compare differentiation in mouse and human ESCs. This will shed light on the differences between the species and take advantage of their different pros and cons: mouse ESCs are easier to engineer, while some human ESCs are able to differentiate into placental cell types without genetic manipulation. The ability to use diverse human ESC lines will be crucial for this. Together, the research proposed here will explore the fundamental basis for decision-making in individual cells. It will thus directly enable better strategies for controlling stem cell differentiation for therapeutic purposes.
Statement of Benefit to California:
Embryonic stem cells have enormous therapeutic potential for a wide variety of diseases. However, a fundamental problem that hinders their usefulness is that they typically give rise to a mixture of many different types of cells, rather than to the particular cell type that is most useful for a given application. The proposed research will analyze the genetic basis for this variability. It will thus play a pivotal role in facilitating the engineering of embryonic stem cells for therapeutic applications that will be of benefit to citizens of California and others.
SYNOPSIS: This proposal focuses on modeling the genetic circuits of pluripotency genes in embryonic stem cells (ESCs), focusing specifically on the well-characterized transcription factors Oct4, Sox2 and Nanog, and the trophectoderm-associated factor Cdx2. Mouse and human ES lines will be generated in which the promoter elements of each of these genes drives distinct fluorescent reporters, and then expression data will be collected on single cells using time-lapse microscopy, under different conditions of cell proliferation and differentiation, including genetic manipulation of expression of each of the factors. The data will be analyzed to develop a mathematical model of the interactions of these genes and their importance for pluripotency and differentiation. IMPACT & SIGNIFICANCE: The impact of this proposal lies in its asking how ESCs give rise to the dynamic and heterogeneous behaviors seen in cell populations, and how these behaviors differ between human embryonic stem cells (hESCs) and mouse embryonic stem cells (mESCs). The proposal is an ambitious one, analyzing pluripotency and differentiation quantitatively and dynamically at the single cell level. The use of automated time lapse microscopy to make movies of cells during growth and differentiation is a potentially useful idea. The use of homologous recombination to integrate fluorescent reporter genes for key transcription factors into their chromosomal loci to facilitate determination of dynamics is proposed. Computational image analysis and cell tracking software will be used for quantitative data analysis. Having such a system available would be important not only for these investigators but for all those involved in the hESC field. Moreover, the idea of comparing regulatory dynamics in the murine and human ESCs to try to understand the “rules” governing their differences in differentiation promises to shed some light on why they behave somewhat differently from one another. Oct4, Sox2 and Nanog are clearly key regulators of ESC pluripotency, but how precisely they function remains unclear. Also, they appear to play different roles in mouse vs. human ESCs. Mathematical modeling of a "pluripotency circuit" would likely provide significant insight into basic mechanisms of ESC biology, and could further generate a predictive model for strategies designed to manipulate ESC function. The studies are innovative and original in that they will generate novel reporter ESC lines, which may be broadly useful to the community, and will analyze gene dynamics in real time and at the single cell level, which has not previously been accomplished. Analysis at the single cell level has the potential to reveal intricacies of gene expression changes that are not appreciable when bulk populations are examined. QUALITY OF THE RESEARCH PLAN: The project will focus on those regulatory systems underlying maintenance of pluripotency and differentiation into trophectoderm lineages for hESCs and mESCs. The PI notes that genes associated with pluripotency (OCT4, SOX2, NANOG) incorporate positive feedback systems which they will analyze by constructing cell lines in which these transcription factors are labeled with three different fluorescent proteins. They will characterize fluctuations and correlations among these components during growth in pluripotent conditions and during differentiation and perturb their activities by ectopic expression of additional copies and use of inducible small hairpin RNA. Data will be analyzed using a mathematical model of the circuit. For trophectoderm differentiation they focus on the mutually inhibitory feedback loop involving Cdx2 and OCT4. They characterize its dynamics under conditions that generate heterogeneous differentiation patterns, and analyze its response to defined genetic and environmental perturbations. Data are analyzed using a mathematical model of the underlying circuit. The research plan is well laid out, relatively straightforward, and addresses an interesting and important question in stem cell research. It synergizes with ongoing research in the applicant's lab, which is primarily focused on understanding gene regulatory networks using individual, live cell imaging technology. Unfortunately, the entire project relies critically on the generation of mouse and human reporter lines that faithfully communicate the expression of each target gene, and the applicant provides no preliminary data addressing the feasibilty or potential efficiency of generating these lines. Also, a modified culture technique may be needed for imaging single cells for these studies. While preliminary data is provided for mESCs, none is provided for hESCs which the applicant argues can be imaged by focusing on cells at the periphery of the colony after plating on fibronectin-coated substrate. However, it not clear that cells at the periphery will behave identically to cells in the center of the colony, and no data is provided, or studies proposed to evaluate this. Also, will local variations in fibronectin density affect cell dynamics, and if so, how will this be assessed? Finally, the applicant states that the temporal sequence of gene expression that is determined from single cell imaging of reporter lines under various conditions will be correlated with cell fate; however, it is unclear how cell fates will be analyzed at the single cell level, or even which cell fate parameters will be monitored. More detail about experimental design and readout is needed here. STRENGTHS: The strength of the proposal lies in adding to our comprehension the comparative biology of the hESC and mESC systems as well as in the potential utility of the techniques developed to other laboratories. The Principal Investigator (PI) does not have significant previous experience with hESCs but will rely on a core facilty directly by M. Pera. WEAKNESSES: The weakness of the proposal lies in the manipulations that must be performed in order to create an environment suitable for making the recordings that are central to the tracking operation. In brief, the cells must be loaded with the appropriate reporter(s) and cultured in monolayer. Either of these conditions in its own right may in itself influence cell development and differentiation. For example, given that cells in the body are largely found in a 3-D rather than a 2-D environment and that those which differentiate into muscle (especially smooth and cardiac) are subjected to contractile forces from early in embryonic development, it is not certain that the evolution of the cells as studied in this monolayer system will be consistent with what happens in a physiologically relevant setting. The result, even if the experiments are brought to fruition, may or may not have meaning to contractile systems and – importantly- may not have meaning to stem cell pluripotency and differentiation into non-contractile cell types. Of some concern is the current state of evolution of preliminary data. In addition, some of the preliminary results presented raise additional questions: for example, in figure 7, it is clear that the two daughter cells appear to be emitting very different fluorescence patterns between 1000 and 1200 minutes. Yet we cannot see what happens next, nor is this difference commented on. An additional concern is the fact that much of the work seems to be driven by the competency of the Core rather than by the PI’s own experience with hESCs. A greater preponderance of preliminary data reflecting the PIs experience here would have been of help. In summary: (1) there are no preliminary data demonstrating that the time-lapse imaging will work for hESCs; (2) the PI has not generated the required knock-in lines, and this will be very challenging, and (3) hESCs are not the primary interest of the applicant's lab DISCUSSION: This proposal aims to track cell fate using time-lapse microscopy and correlate cell fate with changes in gene expression - a straightforward approach. The strength of the proposal is that they can look at single cells. A major concern was that the culture system has to be adapted significantly for microscopy -and culture conditions may bias the data and the results; one reviewer is concerned that proposed conditions (monolayers, etc.) could influence cellular differentiation in non-physiologically relevant ways. Moreoever, only a subset of cells can be followed. The impression of one reviewer is that although this is great technology, hESC work is not the primary interest of the lab, and the PI, an accomplished investigator who has this very interesting technology, will not make a major committment to stem cell biology. The PI will need lots of help from the core facility and lots of reagents, including the tagged protein for the knock-in lines. Another reviewer viewed this as a "your favorite technology" grant, but was swayed by the single-cell imaging which is great and done at a very high level. Unfortunately the PI is attempting to adapt the physiology/structure of their system to match what they want it to do for stem cell research. With respect to the technology, yet another reviewer felt that while the preliminary data were beautiful, the PI can't be certain the evaluation of the cells will be consistent or relevant to what happens in a relevant setting since they will be using cells at the periphery of the colony. These may have very different properties and thus it is a little off center to study them. Figure 7 showing one cell becoming two daughter cells was cited as an example of how this grant seemed a little "off". The preliminary data are beautiful but the pictured cell goes to 2 daughters which go into 2 directions and then the graph ends. What’s the basis for the variation? What does it all mean? At the end of the graph there is text stating that there is variation afterwards, and the whole point of this work is to find out the basis for this variation! There was some disagreement between the reviewers over whether this technology grant should be funded. In summary, the PI is an accomplished investigator in these methods - that is developing a mathematical model using data collected on single cells viewed with time lapse microscopy. This type of study has not been done before with hESCs; the methodology could allow correlation of the expression of pluripotency factors with function. The dynamic evaluation of gene experiments is novel and innovative, and could establish a new approach for ESCs. This approach has been productive in other areas. The strength of this approach is also a weakness. The main concern that the culture system of ESCs needs to be adapted for single cell analysis and this necessitates looking at only a subset of cells in the colony, namely those at the periphery. The PI has no ESC experience and will rely on others. (There is no preliminary data with hESCs. The proposed research will require a lot of reagents to do this study which is unlikely to be a primary focus of the lab. The PI will need to generate a lot of reagents and that could be very challenging. One reviewer was less enthusiastic than the other two because this is an example of adapting a system to match your technology. Either of the modifications that is needed (2D culture environment and gene expression) will change the cells so much that it may be hard to interpret the data obtained. Also, there’s a question of physiological relevance. The problem with working with cells at the periphery of the colony is that they’re different from the ones in the middle.