The Niche

Reprogramming insights: in silico modeling suggests active dedifferentiation

Eric Werner writes:

The recent results of dedifferentiating adult mouse fibroblast cells into stem cells brings into focus the fundamental question of how differentiation and development are controlled. (Cyranoski, D., Nature 447, 618-9; 2007). (Okita, K., et al. Nature doi 10.1038/nature05934, 2007). The fact that just four regulatory genes inserted into cells using viral vectors, can transform normal, differentiated cells into pluripotent stem cells indicates that for some cell types, at least, the process of dedifferentiation is more a process of activation rather than deactivation (Reik, W., Nature 447, 425-32; 2007). Indeed, this falls in line with in silico studies where stem cells, and, more generally, multicellular differentiation and development are modeled on computers.

There the process of dedifferentiation is a process of activating regions of the genome that initiate some developmental process. Since the initial state of the cell early in development of an organism appears to be evolutionarily conserved and phylogenetically ancient, it may on hindsight have been expected that the activating transcription factors of those regions are combinatorially simple in structure. As an organism evolves and requires more discriminating regulatory networks, the concomitant combinatorics of transcriptional activation would need to be more sophisticated. Indeed, it is questionable whether protein transcription factors are sufficient to generated the complexity of regulation and, thereby, the regulatory networks needed for the development of complex organisms (Werner, E., FEBS Letters 579, 1779-1782; 2005). For more complex networks, a hybrid of RNA stabilised with protein helpers agents would seem to be a better solution to the combinatorial problem. It therefore, seems likely that protein transcription factors may just be one facet of dedifferentiation, differentiation and developmental control. To attain full control of stem cell creation via dedifferentiation, computer models suggest that agents with a more specific combinatorial capacity than proteins, will play a key role. Furthermore, once the stem cell is created, activation of particular differentiated cell states, will require that protein transcription factors be supplemented with the more specific binding capacity, and hence, the greater combinatorial power of molecules such as RNA, or, as some have argued, DNA itself (Gaillard, C. et al., J Theor Biol 243, 604-7; 2006). Thus stem cells lie at the heart of development and its control. It is no accident that teratomas observed in these stem cell studies involve both stem cells and differentiated non-proliferating cells. There the dedifferentiation has activated a sub-developmental network which generates the differentiated cells in the teratoma. In computer simulations, the range of cells generated depends entirely on the subnetwork that is activated. The fascinating question is whether this also holds for living systems. If it does, it opens the doors to full control of stem cell creation and differentiation.

Eric Werner

Department of Physiology, Anatomy and Genetics, University of Oxford, Parks Road, Oxford OX1 3PT, UK. Email:


Comments are closed.