First, of course, he needs human ES cells. As Melton explained during a September U.S. Senate committee hearing, only ES cells have "the remarkable capacity to make any kind of cell in the body"—skin, bone, brain, liver or other specialized cells, including pancreas. Because ES cells also reproduce themselves, they could actually become factories for specialized cells to replace those lost through disease or injury.

Millions of patients with conditions such as Alzheimer's disease, Parkinson's disease, cancer, osteoporosis and spinal injuries, as well as diabetes, might benefit from such therapy. "Every family in America has been touched by these diseases and conditions, and now we have the opportunity to offer them real hope," declared Representative James R. Langevin of Rhode Island, a quadriplegic with a damaged spine, when he testified before the same committee. Most scientists believe it will take at least 5 to 10 years, however, to solve the problems involved in translating such hopes into treatments.

From Melton's point of view, the hardest part of the job will be learning how to coax primordial ES cells into becoming just one specific kind of cell—in this case, the beta cells of the pancreas that secrete insulin in response to blood sugar (glucose). Next, the new beta cells will have to be implanted into patients and continue functioning there. Finally, researchers must find ways to prevent a recipient's immune system from destroying the new cells. Only then can they hope "to transplant these cells into diabetics and effectively cure them by keeping their blood sugar under control," he says.

Melton embarked on this quest nearly a decade ago, when his son was diagnosed with type I diabetes—a debilitating disease in which the body's immune system destroys the insulin-producing beta cells. No one can live without insulin, which enables the body to use glucose as a basic fuel. People who cannot make their own insulin are totally dependent on daily injections of it. Melton's son, for example, routinely needs seven blood checks and insulin injections per day to maintain a safe balance between his food intake (which raises the level of glucose in the blood), physical activity (which lowers it) and insulin.

"Many times, particularly when he's playing soccer, we double that number of checks to avoid a crisis," says Melton. Diabetics suffer crises both when their glucose level is too high (this may cause lethargy or unconsciousness and may be life-threatening) and when it is too low (this "insulin shock" develops without warning; it may cause shakiness, confusion, seizures or unconsciousness). They may also face complications such as heart disease, stroke, blindness or kidney failure or require amputation, and their life spans are considerably shorter than average.

Before turning his attention to pancreatic development, Melton had won fame for his work on how the frog's body plan is established early in the life of the embryo. One of his best-known discoveries involved the frog's nervous system; he showed that this most complex part of the body forms simply by default, when a biochemical signal to make skin is lacking. Not surprisingly, his first studies of pancreatic development were carried out in frogs. Then he moved on to mice, which are genetically much closer to humans. Scientists already had accumulated decades of experience with mouse ES cells, and Melton expected to limit his studies to these. The plan changed in the late 1990s, when James A. Thomson of the University of Wisconsin, Madison, and others devised ways to make human ES cells grow in the lab almost as well as mouse ES cells did.

The news of this achievement galvanized Melton. He knew what he had to do: work with human ES cells. But how? Only a few self-perpetuating colonies, or "lines," of human ES cells had been reported in scientific papers, and most of them belonged to private companies that held the patents for them.

At first he collaborated with other scientists on experiments with human ES cells. "This [work] showed that, like mouse ES cells, the human ES cells respond to various growth factors and differentiate," Melton says. "But we could not find a growth factor that made all the ES cells differentiate into a single type of cell. They would differentiate willy-nilly. This implies that we will not find a growth factor, or even a cocktail of factors, that will cause them all to become beta cells. We will need a different method."

Finding this method will require great effort and, most likely, many different ES cell lines, he believes. "I'm especially concerned because we know from mouse work that some ES cell lines are better than others for making endoderm, the embryonic layer from which the pancreas develops. We don't want to take the chance of being restricted to just a few lines, some of which clearly don't grow well."

A New Partnership
At a friend's barbecue about four years ago, Melton, who was then chair of Harvard's department of molecular and cellular biology, met R. Douglas Powers, a professor at Boston College and the scientific and laboratory director for Boston IVF (an in vitro fertilization clinic). Melton told Powers about his work with mouse ES cells and his attempts to learn how these cells make a pancreas. "It's a kind of decision tree," he explained. "We want to know what genes and cells are involved in each decision so we can learn how to direct the cells' differentiation down that pathway."

They also talked about the then-recent discovery that human ES cells could be grown in culture. This led to a discussion of the need for more human ES cell lines. Powers then revealed that his clinic had thousands of frozen preimplantation embryos left over from couples' efforts to produce a pregnancy and that these extra embryos were slated for destruction. When Melton asked Powers whether he would be willing to collaborate with him in using such embryos to produce new human ES cell lines for his research, Powers readily agreed.

Last year, Melton approached Tom Cech, the new president of HHMI, with the proposal he had discussed with Powers. Boston IVF, one of the nation's largest fertility clinics, would supply frozen embryos that were left over from fertility treatments, with the donors' consent. The Boston IVF scientists would then gently thaw these very early embryos, still at the eight-cell stage, and prepare them to be grown in a lab dish.

Then, a few days later, when the embryos have grown into slightly larger blastocysts with a hollow core, Andrew McMahon, newly appointed chair of the department of molecular and cellular biology at Harvard, would do his part. Using his experience in deriving many lines of mouse ES cells, he would tease out some cells from the inner cell mass of the human blastocysts and try to turn them into new lines of self-reproducing human ES cells. This process would take at least six months, including a stage during which the cells' biological characteristics would be identified and confirmed. Finally, Melton would experiment with various combinations of growth factors or other molecules in an effort to prod the ES cells into becoming active beta cells that churn out insulin.

At HHMI, Melton found a receptive audience. "Our primary mission is to carry out the very best in biomedical research, and Doug is one of our researchers," says Cech. "He came to us and said he had the opportunity to do some very exciting and potentially very important research. We evaluated it carefully. Then we decided to fund it, as long as it remained legal and passed review by the Harvard Institutional Review Board. We also entered into an agreement with Harvard and Boston IVF covering the proposal.

"We are comfortable with our decision on all grounds—medical, scientific, ethical," Cech declares. "In fact, considering the potential for human health, we think it would be unethical not to proceed." continued...

Photos: (From top) Kathleen Dooher, Y. Nikas/Stone