Development is a process that single fertilized egg gives rise to organism. In the case of human development, the resulting human body consists of 37 trillion cells with at least 250 different cell types. These cells are generated by sequential cell division and differentiation events regulated in time- and position-dependent manner. All program required for the regulation is written in the sequence of genomic DNA. However, it is still a big mystery how the genome sequence is read out to direct developmental processes, even in the single differentiation event.

The cellular phenotype is defined by the specific gene expression pattern. Human genome encodes about 27,000 protein-coding genes, and different set of the genes expresses in different cell types. The gene expression is primarily regulated by the functions of transcription factors (TFs). There are about 2,000 genes encoding TFs that are divided into two categories, general TFs expressing in any cell types and tissue-specific TFs expressing in particular cell types. Each cell type is characterized by particular set of tissue-specific TFs. The power of these TFs to determine the cell type is demonstrated by the generation of induced pluripotent stem (iPS) cells from somatic cells with 4 TFs (Oct3/4, Sox2, Klf4 and Myc) that express in ES cells. The cell-type-dependent expressions of tissue-specific TFs are also regulated by TFs, indicating that these TFs regulate their expressions each other. To control the differentiation event in time- and position-dependent manner, the expressions of TFs should be regulated by the external signals. Therefore, we can imagine that the TFs form a network which stability is regulated by the external signal integrations (Niwa, Development, 2018). If we find the principle governing the dynamics of TF network, it will allow us to simulate the developmental process based on the genome sequence.

How can we realize such dream? First we need a simple model system of differentiation event that allows easy experimental manipulations to assess the functions of TFs. Mouse embryonic stem (ES) cells is one of ideal models for this purpose. Mouse ES cells continue self-renewal (making daughter cells with same differentiation potential) under the certain signal integration (the leukemia inhibitory factor (LIF) signal in the case of mouse ES cells). They undergo differentiation (losing the stem cell character and become different cell types) by removal of the external signal. The genome engineering in mouse ES cells is easy because of the many techniques established in their long experimental history. Moreover, there is much knowledge about the functions of TFs expressing in ES cells. Among the reprogramming TFs, Oct3/4 and Sox2 are essential for self-renewal because the extinction of one of them results in differentiation of ES cells to trophectoderm in the presence of LIF (Niwa et al, Nat Genet, 2000; Masui et al, Nat Cell Biol, 2007). In contrast, the loss of Klf4 or Myc shows no obvious effect on self-renewal of ES cells (Yamane et al, Development, 2018). In gain-of-function experiments, the artificial maintenance of Klf4 expression supports self-renewal of ES cells in the absence of LIF whereas there is no such effect for Oct3/4, Sox2 and Myc (Niwa et al, Nature, 2009). How such 4 TFs cooperate to induce pluripotent state? We should learn more about the detail functions of TFs expressing in ES cells to solve this puzzle.

We believe that the genetic control of differentiation process by TFs is a pivotal regulatory mechanism. However, to increase the accuracy and regulate the reversibility of the differentiation event, the epigenetic regulation is required in parallel. These two mechanisms should cooperate to drive the differentiation event. However, it is still obscure how they regulate each other for their cooperation. This is also important point to achieve a real modeling of the differentiation event and the developmental process.

To tackle the questions raised above, the following research projects are ongoing.

１．The characterization of the structure of TF network to maintain self-renewal of mouse ES cells

・How is the effect of MEK inhibitor integrates into the TF network activity?

・Molecular mechanism of regulation of the gene expression by multiple TFs

・Characterization of the sub-TF network governing the transient activation of 2-cell specific genes in subpopulation of mouse ES cells

・Molecular mechanism to transmit the TF network activity to daughter cells

２．The analysis of the mechanism for cooperation between TFs and epigenetic regulators

・Molecular mechanism governing the open chromatin structure in ES cells

・Regulatory mechanism of de novo DNA methyltransferase in ES cells

・Molecular mechanism by TFs and epigenetic regulators keep lineage-specific genes poised for future activation or repression in ES cells.

３．Simulation of the differentiation event based on the genome sequence

・Boolean network model of small TF network