Researchers are rushing to create a how-to manual for reprogramming differentiated cells to an embryonic stem cell–like state, and new work shows that processing gene-regulating RNAs known as microRNAs will be an important part of the reprogramming procedure.

So far, four sets of researchers have reported their success in reprogramming human skin cells to behave like embryonic stem (ES) cells (See More induced pluripotency). The trick is to use viruses to carry various genes into the cells. Mostly, scientists have stuck to the 'Yamanaka quartet' — the genes that Shinya Yamanaka of Kyoto University, in Japan, used originally to reprogram mouse cells; however, James Thomson of the University of Wisconsin–Madison tweaked the technique, swapping in other genes known to be expressed in undifferentiated ES cells. Their reprogramming recipe included a gene called Lin-28, which encodes a protein that is known to bind RNA in the cytoplasm.

In a paper published online by Science on 21 February, Srinivas Viswanathan, George Daley, and Richard Gregory from Harvard University, in Cambridge, Massachusetts, describe how they pinned down what this mysterious protein is doing. They observed that a type of unprocessed microRNA (miRNA) disappears in differentiated cells, but not in ES cells, embryonal carcinoma cells and certain tumours1. When they identified the proteins binding to this type of miRNA, they mostly found proteins that were part of common RNA-processing machinery, but they also found the Lin-28 protein. Lin-28 has been implicated in the development of worms, chickens and mice, so this protein seemed like a good candidate to explore. In a series of experiments that increased and decreased levels of functioning Lin-28 in different cell types, the researchers found that certain unprocessed primary miRNA transcripts (in particular, pri-let-7) built up in cells with more Lin-28. In the absence of Lin-28, the miRNAs appeared in their processed form (mature let-7). The experiments indicated that Lin-28 was directly blocking this miRNA processing.

A version of Lin-28 is activated in some liver cancers, and the researchers speculate that processing let-7 miRNA helps “silence the self-renewal machinery,” thus promoting a differentiated state. In fact, independent researchers have reported that breast cancer stem cells have lower levels of let-7 than do cancer cells that do not resist chemotherapy (ref. 2). It would make sense, then, that, blocking this miRNA processing helps cells move from a specialized identity to a less differentiated one, a process that happens both when cancer begins and when cells are reprogrammed in culture.