I study the mechanisms underlying pluripotency and nuclear reprogramming, with the goals of better understanding mammalian development and using this information for regenerative medicine.
During my doctoral and postdoctoral training, I used nuclear transfer as a functional assay for the reversibility of differentiation and cancer. For example, I was able to show that mature lymphocytes and melanoma cells remain amenable to reprogramming into a pluripotent state, thus showing that the epigenetic changes that take place during terminal differentiation and tumorigenesis are reversible. In an attempt to reprogram adult cells directly, I developed a mouse model in which the embryonic transcription factor Oct4 could be activated; although Oct4 expression alone was insufficient to reprogram cells to pluripotency, I found that adult stem cells were blocked in their differentiation and formed tumors, indicating striking similarities between adult and embryonic stem cells.
In 2006, Shinya Yamanaka and colleagues (Institute for Frontier Medical Sciences, Kyoto University) discovered that the combined expression of four transcription factors—Oct4, Sox2, Klf4, and c-Myc—is sufficient to convert somatic cells into embryonic stem (ES) cell-like cells, called induced pluripotent stem (iPS) cells. However, these initially reported iPS cells showed only partial epigenetic reprogramming and failed to produce viable mice. In my own laboratory, we have shown that identifying iPS cells based on the pluripotency markers Oct4 or Nanog gives rise to pluripotent cells that are indistinguishable from ES cells. For example, iPS cells identified by these markers reactivated the silent X chromosome in female somatic cells and showed global DNA and histone methylation patterns that are highly similar to ES cells. Moreover, these iPS cells gave rise to chimeras that contribute to the germline, thus fulfilling several criteria of pluripotency.
Using doxycycline-dependent lentiviral vectors for the four factors, we showed that transgene expression is required for 8–10 days before cells reactivate their endogenous pluripotency program and thus become stably reprogrammed. During this time, cells initially down-regulate somatic markers before activating embryonic genes and eventually turning on telomerase and the pluripotency genes Oct4, Sox2, and Nanog. Thus, reprogramming follows a defined sequence of molecular events that can be traced with specific markers; this will allow us to further dissect the mechanism of induced pluripotency.
The generation of iPS cells is inefficient (0.01–0.2 percent), raising questions about the barriers inherent to nuclear reprogramming. Over the past couple of years, we have tested several possibilities that may explain this. One possibility for the low efficiency has been that iPS cells can only be generated from undifferentiated adult stem cells, as these cells may be more amenable to reprogramming than terminally differentiated cells. We have excluded this possibility by using the four initially reported transcription factors to generate iPS cells from defined, genetically marked cell populations—including pancreatic β cells, melanocytes, and B and T lymphocytes. However, the notion that these mature cells can be reprogrammed into iPS cells did not rule out that adult stem or progenitor cells may be converted into iPS cells more easily than differentiated cells. We have recently demonstrated that hematopoietic stem and progenitor cells yield iPS cell colonies several hundredfold more efficiently than mature blood cell types, suggesting that their epigenetic state is more susceptible for reprogramming than that of differentiated cells. We are trying to understand which aspects of stem/progenitor cells make them more amenable to reprogramming.
Another possibility for the low efficiency of reprogramming has been the low infection efficiency of somatic cells by viruses expressing all four reprogramming factors at appropriate levels. We have developed "secondary systems" to activate the four transgenes homogeneously in all somatic cells. This resulted in the formation of iPS cells at increased efficiencies of 1–5 percent, indicating that the low efficiency of viral infection contributes to the low reprogramming efficiency but is not the only limiting factor.
Activation of endogenous genes by insertional mutagenesis has been discussed as a potential requirement for the generation of iPS cells, which might also be the cause for the low efficiency. In collaboration with Thomas Graf's lab (Center for Genomic Regulation, Barcelona), we have mapped viral insertion sites of several iPS cell lines, and we did not find commonly targeted genes. This suggests that activation of a single endogenous gene through insertional mutagenesis is unlikely. To provide proof that insertional mutagenesis is not necessary for the formation of iPS cells, we generated iPS cells using adenoviruses, which transiently delivered the reprogramming factors inside the cells without integrating into the genome. The demonstration that iPS cells can be generated without the use of persistent transgenes eliminated a major roadblock for the potential clinical application of iPS technology.
Based on these observations and results from other laboratories, we surmise that stochastic epigenetic events are required in addition to expression of the four factors; this limits the efficiency of reprogramming into iPS cells. To identify these stochastic events, we recently analyzed the role of cell cycle–dependent kinase inhibitors, such as p16/p19 and p53, during the reprogramming process. We discovered that down-regulation of the p16/p19 inhibitors is critical during the reprogramming process to ensure derivation of pluripotent, immortal cell lines. Fibroblasts deficient for these cell cycle regulators give rise to iPS cells with a faster kinetics and at higher efficiency than wild-type cells, indicating that silencing of this pathway is one of the roadblocks during the reprogramming of normal cells. Transient inhibition of cell cycle inhibitors during human somatic cell reprogramming may thus be a way to generate patient-specific iPS cells efficiently.
My current efforts are aimed at identifying additional modulators of the reprogramming process to identify molecules whose manipulation can aid in the generation of iPS cells. Specifically, we are performing RNAi screens to identify activators and suppressors of reprogramming. We are also establishing epigenetic and transcriptional profiles of the intermediate stages of reprogramming, with the goal of establishing the sequence of molecular events necessary for a somatic cell to assume a pluripotent state. Finally, to address whether iPS and ES cells are equivalent, we are developing transgenic tools to compare their molecular and functional properties. We are using murine and human cellular systems to test the role of novel genes in pluripotency, reprogramming, and development.
Grants from the National Institutes of Health, the Sydney Kimmel Foundation for Cancer Research, the V Foundation for Cancer Research, and the Harvard Stem Cell Institute provided partial support for these projects.
As of May 30, 2012