Ebola protein explains deadly mystery.

A gruesomely detailed account of a 1989 Ebola virus outbreak in a monkey house in Reston, Va., terrified millions of readers of Richard Preston's The Hot Zone (1994, Random House). Ebola, normally found in Africa, kills people with chilling efficiency most of the time. Yet the strain that surfaced in Reston spared workers who became infected even as it slew monkeys.

A research team based at the National Institutes of Health in Bethesda, Md., may now have uncovered the key mechanism by which Ebola destroys cells in blood vessels and elsewhere, bringing about massive bleeding and other symptoms. The scientists may also have unraveled why Ebola didn't prove fatal to people in the Reston facility.

"This definitely puts a potential explanation on the table," says Gary J. Nabel, whose team at NIH describes its findings in the August NATURE MEDICINE.

Two years ago, Nabel, Zhi-yong Yang, and their NIH colleagues reported their initial studies of Ebola's glycoprotein. The virus produces this sugar-laden molecule in a secreted form and in a larger version that becomes part of the surface of the virus' new copies. The secreted form appears to suppress immune cells, whereas the surface glycoprotein binds to so-called endothelial cells, which make up the lining of blood vessels (SN: 2/14/98, p. 102).

Nabel and his colleagues found that the surface glycoprotein does more than help Ebola home in on endothelial cells. When the scientists added the gene encoding the protein to laboratory-grown human cells, they observed that the cells assumed a rounder form over the next 24 hours. The cells also separated from one another and, 2 to 3 days later, died.

These observations indicate that it's not Ebola's binding to a cell that triggers its death but the later synthesis of viral glycoprotein inside the cell that proves fatal. Nabel suggests the Ebola protein must build up to a certain threshold before it kills. This would give the virus enough time to force an infected cell to make new copies of the virus. The cell's death then would spread those viruses.

In addition to the simple cell experiments, Nabel's team worked with intact blood vessels taken from people and animals. The researchers infected those cells with a cold virus they had engineered to carry the Ebola glycoprotein gene. Within 48 hours, massive numbers of endothelial cells began to die and the blood vessels became leaky. Such effects could lead to the internal and external bleeding caused by Ebola.

"I think this accounts for some aspects of the hemorrhaging. It's likely this is a starting point of a cascade of events," says Nabel. "This creates havoc in the sense that it causes a lot of cell death that then triggers clotting abnormalities and other problems."

The use of human blood vessels in the experiments makes Nabel's conclusions particularly compelling, says Robert A. Lamb, who studies viral glycoproteins at Northwestern University in Evanston, Ill.

By mutating the glycoprotein gene to produce variants that don't kill cells, the investigators also identified a region on the viral molecule that's necessary for its destructive behavior. Previous research has shown, says Nabel, that the Reston strain of Ebola differs in this region when compared with the killer strain from Zaire that his group investigates.

Indeed, when the NIH scientists repeated their experiments with the glycoprotein from the Reston strain, the molecule killed cells and destroyed blood vessels of nonhuman primates but didn't harm human cells.

The next major task for Nabel's group, says Lamb, is to ascertain how Ebola glycoprotein kills human cells. Scientists suspect that it binds to a protein inside the cells, and Nabel speculates that drugs thwarting this interaction would limit Ebola's spread in the body and give the immune system time to clear an infection.