The standard technique for creating make differentiated cells behave like embryonic stem cells uses viruses to insert the genes cMyc, Klf4, Oct4, and Sox2 into cells, but adding these genes to cells makes them less predictable and more likely to form tumors. Researchers have been able to reprogram neural stem cell using only Oct4, but these cells are not readily available from patient biopsies and so researchers are searching for alternate techniques. New work published in Cell Stem Cell shows that a small druglike molecule can effectively replace two of the four genes typically used to generate induced pluriptotent stem cells.
To begin their hunt for compounds that could help reprogram cells, researchers led by Kevin Eggan and Lee Rubin of the Harvard Stem Cell Institute used cultures of mouse skin cells engineered to express green fluorescent protein as a marker of pluripotency. They first screened for small molecules that allowed mouse cells to be reprogrammed without adding the gene for Sox2. When three such molecules were identified, the researchers tried again and found that one of the molecules could reprogram cells even in the absence of cMyc, a tumour-promoting gene that, while not required for reprogramming, greatly boosts reprogramming rates.
To make sure the cells were really reprogrammed, the researchers performed a series of tests, including mixing them with mouse embryos and demonstrating that they could contribute to every type of tissue in chimeric mice. They named the identified molecule RepSox for its ability to replace Sox2 and also after the Red Sox, the local baseball team. Previous studies had identified this molecule as inhibiting a pathway known as TGF-beta signaling. Careful work showed that RepSox did not work by activating the Sox2 gene in fibroblasts, as might be expected. Instead, the molecule functions in partially reprogrammed cells that accumulate in the absence of Sox2, apparently by inducing and stabilizing Nanog expression. Thus, the researchers write, the discovery of RepSox is important not only for replacing one of the reprogramming factors but for illuminating a new strategy to identifying such molecules. “There need not always be a discrete, one-to-one mapping between the functions of the reprogramming factors and their chemical replacements.”
Robert Blelloch, who studies reprogramming at the University of California San Francisco, praised the team’s strategy of only screening compounds whose mechanisms are at least partly understood. “They find a small molecule that replaces a factor, but they take it further and use it to understand the biology.”