Who Am I: Blurring the Lines of Cellular Identity
- Author(s): Ho, Chi Kent;
- Advisor(s): Sun, Yi E;
- Teitell, Michael A
- et al.
Cell identity used to be viewed as something static and each generation of organisms had their primordial germ cells separated early so that the somatic cells are fated to differentiate into the rest of the body. Pluripotency was a one-way street where things went from pluripotent to unipotent. However, today's understanding of cell identity has been massively shifted as the field of reprogramming has matured to a point where changing from one cell type to another is possible.
Attempts at dissecting the process of iPS reprogramming has helped to improve efficiency of reprogramming and lower tumorigenicity, but improving reprogramming speed proves to be a stronger barrier. And yet, many different cell types reprogram into iPS faster than with fibroblasts. By analyzing and comparing the gene expression profiles of two human cell types that reprogram quickly (Amniotic-Fluid Derived Cells and Adipose Derived Stem Cells) and two human cell types that reprogram slowly (Human Foreskin Fibroblasts and keratinocytes), I found two sets of genes that could explain the differences in reprogramming speeds. PIAS3 and STAT3 seem to be the most promising candidates, but experiments with these two factors are inconclusive.
There has also been major progress in direct reprogramming or transdifferentiation between two lineage-specified cell types. One of the first cell types to be successfully transdifferentiated into in this new revitalization of reprogramming was the induced neurons. However, the neurobiology field is no longer dominated solely by the neuron as astrocytes have become more important in disease pathology as well. Therefore, I use the tried and true reprogramming approach that successfully reprogrammed fibroblasts into iPSCs and iNs to reprogram mouse and human fibroblasts into functional induced astrocytes. Mouse iAs require only the transcription factor, NFIA while human iA reprogramming requires NFIA in addition to TET2 and NOTCH. As a proof-of-principle, I apply the technology to reprogram Alexander's Disease patient-specific fibroblasts to show that the induced astrocytes can be used to create disease models in vitro. Finally, I use shRNA knockdown of the mutant protein that causes the mutant phenotype to show that the iA model is tractable to therapeutic testing.