The blood hormone thrombopoeitin,
or TPO, has been called the rarest thing on earth. The
existence of this hormone, which stimulates the production of
platelets, the blood's clotting agents, had been hypothesized
since about 1960. But if it existed, it was present in blood at
such low concentrations as to make finding it impossible.
Scientists compared looking for it to "chasing a 
Isolating the hormone had been a quest of scientists for
more than 30 years, but no one had succeeded; some researchers
even speculated it did not exist. But a discovery in 1986
touched off a new race to find TPO. The French scientist
Francoise Wendling discovered a receptora protein protruding
from the surface of cells that can latch onto specific target
molecules, as a lock fits a key. It appeared to be a growth
factor receptor related to blood cells. No one was sure what it
did, so initially it was dubbed the "orphan receptor." Could
this be the "lock" that fits the TPO "key"? As many as six to
eight teams from corporate and academic laboratories were
scrambling to be the first to find out.
By July of 1994, four groups had reported, nearly simultaneously, the isolation of TPO. A new contest for the patent had begun.
One of those groups was a team working at Seattle-based ZymoGenetics company, led by Donald Foster, in collaboration with Kenneth Kaushansky, UW professor of medicine. They developed a unique and unconventional approach to isolate TPO.
Competitors had cloned the receptor and fashioned a sort of molecular "filter" out of it: They attached many of the receptors to the surfaces inside a tube, and then passed blood through the tube, with the strategy that the receptors would "catch"and concentrate TPO molecules as the blood passed by. Bay-area basedGenentech Corporation succeeded in isolating approximately two-millionths of a gram of the hormone in this way.
The UW-ZymoGenetics team took a completely different tack. They took a type of mouse leukemia cell that grows only if it is stimulated by hormones, and they engineered it so that the hormone it would respond to is the hormone that binds to the orphan receptor. Next, the researchers mutated the engineered mouse cells in an attempt to activate the gene that codes for TPO. Any one mutation might not strike at the right spot to activate the TPO gene, which is normally turned off. But if enough cells are used, the chances are that the right spot will be activated in at least one of them. Only those cells that mutated to turn on the right gene would survive and grow. The gene for TPO could then be identified, cloned, and used to produce useful quantities of the hormone.
The team began by exposing 10 million mouse cells to an agent that causes mutations. And sure enough: one mouse cell lived that was making TPO. Success of the experiment was confirmed by isolating a tiny quantity of the hormone and administering some to mice, whose platelet production jumped up by about 400 percent.
Platelets are produced when cells called megakaryocytes, large bone-marrow cells, swell and burst into about a thousand fragments: the platelets. TPO causes immature bone marrow cells to become megakaryocytes and stimulates megakaryocytes to produce platelets. Kaushansky has found that administering extra TPO to an animal increases the number of magakaryocytes tenfold, and also stimulates each megakaryocyte on average to develop 2,000 to 3,000 more platelets than usual.
Genetically engineered TPO holds great promise as a treatment for blood clotting disorders. Patients receiving chemotherapy or radiation treatments for cancer, bone-marrow transplant patients, children with certain viral diseases, and certain AIDS patients, all of whom may suffer from a shortage of platelets, may be helped in the future with TPO.
