26 Aug 2005

For now, embryonic stem (ES) cells remain perhaps the best hope for creating replacement cells and tissues that could one day be used to treat a myriad of diseases, including neurodegenerative disorders like Alzheimer and Parkinson diseases. But human embryonic stem cells remain controversial because the embryos from which they are derived are destroyed in the process. The same ethical dilemma shrouds therapeutic cloning, or somatic cell nuclear transfer, a process for deriving cells and tissues that are a genetic match to the intended recipient, because this process also involves formation and destruction of embryos. But what if there was a way to generate totipotent, tailor-made human cells without having to create new embryos? In this week's Science, Kevin Eggan and colleagues at Harvard University report a method that may have some future in this regard.

During somatic cell nuclear transfer, the nucleus from a donor cell, often a skin fibroblast, is placed inside an enucleated oocyte. The cytoplasm of the oocyte "reprograms" the genetic material in the nucleus, effectively setting it back to an embryonic state. But what if this process were reversed by inserting the reprogramming machinery into the fibroblast? Might this not reprogram the fibroblast directly, turning it into a stem cell while bypassing embryogenesis? In fact, researchers in Japan (see Tada et al., 2001) have already shown that the technique works for mouse cells.

Now, first author Chad Cowan and colleagues have found that it works with human cells, too. To test the idea, the authors fused human fibroblasts with human ES cells. They then examined the hybrid cells to see if the adult cell nucleus had indeed been reprogrammed. They found that the hybrids assumed many characteristics of stem cells. For example, they grew in compact tight clusters, just like stem cells, and they could differentiate into many cell types, including neurons (cells expressed neural-specific tubulin), muscle (expressing muscle myosin heavy chain), and intestinal endodermal cells (expressing α fetal protein).

Cowan and colleagues used a variety of approaches to prove that the stem cell properties of these hybrids derived from the fibroblast chromosomes and not those contributed by the stem cells. A green fluorescent protein chimera of the Rex1 embryonic gene that was inserted into the fibroblast genome was turned on once the cells fused. But more importantly, transcriptional profiling showed that of over 6,000 genes in the fibroblast genome that are either quiescent (2,541 genes) or overexpressed (3,867 genes) compared to ES cells, transcription of the vast majority was reset to ES-like levels after fusion. Only 29 genes retained their low level of expression and only 12 remained more highly expressed than in ES cells. The results suggest that over 99 percent of the genes in the somatic cell genome had been reprogrammed to ES-like transcriptional levels. Some of these genes, such as Oct4 and Nanog (see ARF related news story), are known markers of cell pluripotency.

These results suggest that the genome of adult cells can be reprogrammed in situ, without removing the nucleus from the adult cell. The advantage from an ethical standpoint is that existing stem cells could be used to generate new tailor-made stem cells without the formation of new embryos. "Eventually, this approach might lead to an alternative route for creating genetically tailored hES cell lines for use in the study and treatment of human disease," write the authors. Before this can happen, however, there are many technical hurdles to overcome. For one, the extra set of chromosomes derived from the ES cells would need to be removed. What may prove more effective is fusion of enucleated ES cells, that is, cytoplasts, with adult cells. However, though this technique has been attempted before (see, for example, Do and Scholer, 2004), it has not yet proven successful.—Tom Fagan