“The general concept that there are classes of transcription factors that yield paradoxical results—both activating and inactivating pathways—has been around,” says Epstein. “What's an addition here is a molecular mechanism.”
Epstein's group came across the phenomenon while studying neural crest cells. Pax3 was required in mice for the eventual induction of melanocyte differentiation markers such as dopachrome tautomerase (Dct), but in cell culture experiments Pax3 was found to repress Dct expression.
The paradox could be understood based on a series of relationships. Epstein found that Pax3 helped turn on Mitf, and Mitf was needed to turn on Dct. But if Pax3 was still around, it could prevent Mitf from binding and turning on the Dct enhancer. Only when an activated Wnt signal displaced Pax3 was Mitf able to do its job.
Consistent with this model, only areas positive for Wnt signals were expressing both Pax3 and Dct. In the embryo, such Wnt-expressing areas are thought to be terminal migratory locations for neural crest cells. In the adult, skin damage may generate Wnt signals that prompt the necessary melanocyte differentiation.
The group suggests that Mitf acts like a “biological capacitor.” It builds up, but is repressed until needed. Epstein is now looking for a similar organization in other stem cell systems. “It's a tight little circuit,” he says, “and I bet it will come up again and again.”