How Cells Listen to Many Voices-and Get Them Right / New Class of Signaling Pathway Gives Clues to Growth Factor's Powerful Effects

Like the hormone estrogen, transforming growth factor beta (TGFb) is a multifarious but Janus-faced molecule: Not only does it perform different functions in different tissues, it also is capable of triggering opposite effects. TGFb and other members of this so-called superfamily of growth factors are able to spur the proliferation of cells, but also appear to suppress tumors; they close wounds by generating and structuring new tissue, but also clog up and stifle organs with unnecessary scar tissue Moreover, TGFbs can manipulate how cells react to other growth factors. And in embryonic development, they are involved in most of the crucial steps that allow the early embryo to unfold properly, such as setting up the future body axes and inducing the germ layers from which all organs then arise. Finally, TGFbs act in different ways at different developmental stages, inducing formation of a tissue at one stage and, just days later, telling individual cells whether to become a neuron or one of the brain's glial support cells.

For 20 years, researchers were perplexed about how one family of related proteins can pull off such a bewildering array of feats in different tissues of the body. How do the TGFbs convey a message encoding each distinct function without the information getting scrambled? How are the receiving cells to know just which of TGFb's many commands to carry out? "We really know nothing about how the extracellular TGFb signals actually translate into individual patterns of cell-type-specific gene expression," says Malcolm Whitman, assistant professor of cell biology.

Now, Whitman and two coworkers have found the first answer for any member of this family. The factor they studied, activin, binds to cells of the very early frog embryo and With Whitman's finding, a chain of signaling steps falls into place. Several research teams, who had been carving out the pathway from different ends, are suddenly hitting on each other's work, like crews digging a tunnel from different sides of the mountain. The membrane receptors binding the TGFb have long been known, but just last year, Bill Gelbart's team at Harvard University identified the intermediary messengers, dubbed MAD proteins. Other researchers discovered soon afterwards that the MAD proteins acted like shuttle diplomats: They linger in the cytoplasm poised for a signal and, after receiving it, move to the nucleus, supposedly to regulate gene transcription. Whitman's finding closes that chain.

Taken together, these studies introduce "a new class of signaling pathway," says Whitman. Given TGFb's huge range of effects, it is "a surprisingly simple one." Contrary to the long bucket brigades of proteins required in previously known pathways, the TGFb pathway uses only one protein But if the pathway is simple, the original conundrum seems to raise its head again: What mechanism makes it possible to keep all of TGFb's messages from being garbled? Most important in answering that question will likely be DNA-binding proteins like FAST, says Whitman, which may occur in different tissues in distinct patterns. They may guide the incoming MADs to a highly select group of genes While it is too early to tell if this hypothesis is right, Whitman's first example of the workings of such a nuclear factor will give researchers a handle on finding out. And understanding tissue specificity is just what they need to do in order to get a handle on TGFb's subtle roles in cancer and other diseases. That is because blocking only certain aspects of the growth factor's action could treat diseases specifically, whereas knocking out all of TGFb1s effects could well do more harm than good.

The authors of the Nature article, in addition to Whitman, are graduate student Xin Chen and postdoctoral fellow Melissa Rubock, both in the Department of Cell Biology at Harvard Medical School.