Identification of Regulators of Neuronal Differentiation by High Throughput Screening of Embryonic Stem Cells

Murine embryonic stem (ES) cells were first derived over twenty years ago from cultured blastocysts and have been primarily used as a tool to better understand mouse development through targeted gene deletion and promoter/enhancer trapping. Only recently, the focus has shifted to stem cells and the molecular mechanisms by which they choose a specific cell fate. Interest in signaling and differentiation mechanisms in stem cells and directing stem cells has gained momentum because of the success (albeit limited) in animal models of disease transplanted with stem cells or their differentiated derivatives. Most of what is known about individual protein's roles in stem cell differentiation comes from differentiating ES cells in defined medium and/or on defined substrates and observing changes in cell morphology and protein marker levels. Similar assays performed with knockout ES cell strains have, to an even greater degree, also contributed to our fundamental understanding of the role of particular proteins in differentiation. Though such studies have been the basis for proving ES cells' totipotentcy, they often result in mixed populations of cells making it difficult to correlate a given protein's expression with the differentiation of a particular cell type. The problem of mixed cell types has been partially circumvented through the use of cell lineage specific promoters driving the expression of a drug resistance gene such as puromycin or a fluorescent reporter such as green fluorescent protein (GFP). Even with the ability to select cell lineage, there has only recently been an effort to combine this approach with overexpression studies to show that an individual protein's presence has the ability to direct the differentiation of ES cells into a specific cell type. For many labs, the handful of screening methodologies that have been published are not feasible because of the cost involved, and/or are not suited for many complex multistep assays. Transcription factors have been shown to be an integral part of the ES cell differentiation process. Deletion or overexpression of single transcription factors has been shown to independently direct differentiation and dedifferentiation of stem cells or post-mitotic cells, respectively. A prime example of 'forced' differentiation would be converting various cell lines, including ES cells, into myotubes simply by expressing the transcription factor MyoD. Adding to the number of such associations would greatly increase our basic understanding of many newly cloned transcription factors with unknown functions. While the ability to perform mammalian cell-based assays with a large number of cDNAs is not a novel concept, the novelty about the approach of 'one gene per well' is that a specific, complete cDNA can be directly associated with a given phenotype with much greater speed and accuracy. The major limitation of single clone assays is that isolating individual full-length clones from RNA by RT-PCR or direct PCR screening of cDNA libraries is a laborious process. Fortunately this is no longer a concern because a number of full-length sequenced clones are now freely available as part of the Mammalian Genome Collection (MGC) effort at the National Institutes of Health, or as commercially available libraries that have been directly derived from the MGC collection. The goal of my thesis has been to examine the role of both well characterized and uncharacterized transcription factors in converting ES cells into neurons. Using the commercially available Gateway system and Human ORFeome 1.1, an unbiased collection of stable mES lines expressing full-length human transcription factors was assembled....