For the past six years, Mark T. Keating has been trying to figure out why humans and other mammals who lose a limb or an organ can't do what salamanders, worms and fish do in such cases: grow a new one. In mammals, cells that try to repair an injury just form a layer of useless scar tissue over it. But in newts, the cells that swarm over a fresh wound rapidly turn into a layer of stem cells that effect a complete repair.

Keating, an HHMI investigator at Children's Hospital in Boston, now believes this is the standard way that many animals regrow lost or damaged body parts. First, the wounds stimulate some mature cells to revert to their infancy as primitive stem cells, or "dedifferentiate"; then these cells are reprogrammed to form new tissue or organs. Certain genes are turned on to start both operations. According to Keating, "humans probably still have the genes that enable other animals to regenerate tissue, but these genes were silenced, turned off" during evolution. Therefore, they might be turned on again, given the right stimulus.

Keating and his colleagues identified their first such gene, msx-1, several years ago while doing research on human heart disease at the University of Utah. They discovered that msx-1 turns on in newts whenever these animals need a replacement part. Mice have a similar msx-1 gene, and the scientists found a way to turn on this gene at will in mouse muscle cells.

Next, Keating, who had moved to Children's and Harvard Medical School by then, used a different stimulus—an extract of regenerating limb tissue collected from newts after their forelimbs had been amputated. This extract worked wonders on mouse muscle cells, making about 18 percent of them dedifferentiate and reenter the cell cycle—not very much less than the 25 percent of newt muscle cells that did the same. In their report on this work, Christopher J. McGann and Shannon J. Odelberg (both at the University of Utah) and Keating wrote that these studies "indicate that mammalian cells have retained the intracellular signaling pathways required for de-differentiation" and that the primary obstacle to regeneration in mammals may be the lack of signals to start the process.

More recently, Keating and his associates compared how zebrafish and mammals react to heart injury. They found that zebrafish regenerate heart muscle with little scarring. However, other vertebrates respond by producing large connective-tissue scars. The researchers then proposed that "scarring and regeneration compete" and that the vigor of each process is critical. In normal zebrafish with heart injury, for instance, a fibrin clot formed, but cardiac muscle fibers "invariably penetrated the clot and constructed a bridge of new muscle around the wound." By contrast, zebrafish with a mutation that impairs cell division produced fewer muscle cells and ended up with big scars. The researchers now believe that stimulating heart muscle cells to proliferate, in response to the proper genetic signals, "will reduce scar formation and facilitate cardiac regeneration in mammals as well." Keating adds, however, that "it'll be a while before anything of this sort is tried in humans."

—Maya Pines

Download this story in Acrobat PDF format.
(requires Acrobat Reader)

Reprinted from the HHMI Bulletin,
December 2002, pages 22-27.
©2002 Howard Hughes Medical Institute