In a striking demonstration of cellular flexibility, scientists have created functioning neurons from fibroblasts, without going through an intermediate pluripotent stage, according to a study published online this week in Nature.

Mouse cingulate cortex neurons Image: Wikimedia commons

"It's really exciting," said molecular geneticist Mathias Treier of the European Molecular Biology Laboratory and the University of Cologne in Germany, who was not involved in the research. "It shows that cells can switch their fate" without going through the pluripotent state, avoiding the potential for tumor formation. "[In] the future, with the right cocktail mix, this [might be] possible for other tissues and organs," he added.

Inspired by Shinya Yamanaka's discovery of four transcription factors that could induce a differentiated cell to regress to a pluripotent state, which could then be redifferentiated into another adult cell type, stem cell biologist Marius Wernig of Stanford University School of Medicine wondered if there might be a shortcut. Rather than going backward to go forward, he wondered, "can we turn a skin fibroblast directly into a neuron?"

To answer this question, Wernig and his colleagues injected mouse embryonic fibroblasts (MEFs) with lentiviruses containing 19 genes that are expressed in neural tissues. Sure enough, 32 days later, the MEFs displayed typical neuronal morphologies. By testing various 5-gene and 3-gene sets of the 19 original factors, the researchers narrowed the field down to just three genes -- Ascl1, Brn2, and Myt1l -- that could convert not only MEFs to neurons, but postnatal fibroblasts as well.

This is not the first example of conversion from one adult cell type to another. In 2008, for example, scientists successfully converted mature pancreatic exocrine cells into cells resembling β-cells in adult mice. More recently, a study reported that the loss of a single gene, Foxl2, induced the conversion of ovary into testis in adult mice. These studies support the idea that a "pluripotent stage may not be essential for the transdifferentation between terminally differentiated cells," molecular biologist Xiangru Xu of Yale University, who did not participate in the research, wrote in an email to The Scientist.

Several studies have also converted a variety of cell types into neuron-like cells. The new report, however, is the first to create fully functional neurons -- capable of generating action potentials and forming synaptic contacts with other neurons -- from another somatic lineage, Wernig said. "When you grow [these induced neuronal (iN) cells] on a preexisting neuronal culture, they are able to functionally integrate with this preexisting neuronal network," he said. "Those are true derived neurons."

A major advantage to skipping the pluripotent stage is the avoidance of tumor formation, Wernig said. Because pluripotent cells are highly proliferative, "trace amounts of these cells in the graft can explode after you transfer these cells into the brain and form a specific type of tumor called [a] teratoma," he explained. "In contrast, these factors [used to form neurons] are really inducing the cells to stop cell division. They're almost anticancer genes."

The disadvantage to this property of the iN cells, however, is the inability to expand them in culture, Treier said. While the pluripotent state allows for the production of large numbers of cells, with this method it's "only what you get out of a patient [that] you can convert," he said. "There's really a one to one conversion."

Thus, despite the fact that the researchers achieved nearly 20% efficiency in the conversion -- well above the 0.1% or 0.01% efficiency associated with inducing pluripotency -- generating enough tissue to be useful in a therapy is unlikely, said Sheng Ding of the Scripps Research Institute in La Jolla, Calif, who did not participate in the research. "That's going to be the challenge to [using] this type of technique in the future."

Furthermore, to induce the cellular change the researchers are still using genetic manipulation, a method that is inherently risky. Even without using oncogenes, inserting factors into the genome can be potentially harmful if, for example, the manipulation activates a native oncogene or disrupts the function of a tumor suppressor gene. "You have to be very careful in terms of those genetic manipulations in cells," Ding said.

This problem can be overcome by replacing the genetic factors with nongenetic manipulations, Wernig pointed out, such as protein transduction, nonintegrating lentiviruses, or plasmids. But for now, he added, "a very attractive application" of this method could be the development of specific disease models. "It's not only the neurodegenerative diseases, like Parkinson's Disease," he said, "but also diseases which affect the neuronal activity of the brain, such as depression or schizophrenia or even autistic disorders."

"This is very exciting work and likely the beginning of a wave of experiments growing out of the Yamanaka breakthrough," stem cell biologist George Daley of the Harvard Stem Cell Institute and Children's Hospital Boston wrote in an email to The Scientist. "This is only the tip of the iceberg," Treier agreed. "I think we will see many more examples in the near future."


Related stories:

One gene keeps ovaries female
[10th December 2009]

Differentiating Hope from Embryonic Stem Cells
[15th December 2003]

Short cut to neuron differentiation
[13th August 2002]