Genetically engineered fungus fights blight.

Once a dominant tree in eastern North America, the mighty American chestnut was felled by a fungus introduced from Asia at the turn of the century Now, molecular biologists have developed a strategy for disarming this fungus so that a new generation of chestnuts may one day tower in the forest.

The strategy improves upon the use of a less deadly strain of chestnut blight to neutralize killer strains. Rather than destroy bark and make the tree wilt and die, this "hypovirulent" strain causes only superficial, temporary sores on the bark, says Donald L. Nuss of the Roche Institute of Molecular Biology in Nutley, N.J.

A viral infection reduces this fungal strains ability to destroy the tree, Nuss and Roche colleague Gil H. Choi report in the Aug. 7 SCIENCE. By making DNA that encodes the virus' RNA, Nuss and Choi plan to harness this virus - or an improved version of it - for controlling chestnut blight.

"It's a new and novel approach for a pathogen that's devastating:' comments James L. White, a biotechnologist with the U.S. Department of Agriculture in Hyattsville, Md. "For fungal biocontrol, [this strategy] may be very important."

For more than a decade, plant pathologists have recognized that the less deadly chestnut blight contains double-stranded RNA - a virus of sorts - in its cells. Nuss and Choi proved that this virus renders the fungus hypovirulent.

They began by piecing together a gene for the virus' RNA. When they transferred that gene to virulent fungus, the fungus underwent a transformation: Like the hypovirulent strain, it made less orange pigment and less of certain enzymes. The transformed fungus also caused small cankers to develop on a chestnut stem rather than large, rapidly expanding ones, says Nuss.

When they examined the fungal tissue, Nuss and Choi discovered that the gene did lead to the production of viral RNA.

Some plant pathologists have treated blighted chestnuts with naturally hypovirulent fungus. That fungus sends out threads that merge with the blight fungus, infecting it with the virus and making it less damaging. But in North America, the fungi are often too different for their tissues to fuse, so the treatment fails.

"We're introducing the virus in a new way," Nuss says. The scientists plan to spray spores from the genetically altered hypovirulent fungi onto infected trees.

Now that the virus' genetic information is transferred along with fungal DNA during sexual reproduction, "we can introduce the virus into any strain," says Bradley I. Hillman, a plant virologist at Rutgers University in New Brunswick, N.J. "It effectively expands the range of the virus [so it infects more strains]."

Next, Nuss and Choi plan to study whether modifying the fungus changes what species of tree it will attack. Then they will apply to the U.S. Department of Agriculture for permission to treat blighted chestnuts in field tests. Other researchers have experimented with altered viruses for insect control, but the new tests would represent the first use of bioengineering to harness a virus to control a fungus, says White.

Nuss and Choi are also modifying this gene to improve the virus' ability to disarm the fungus. In addition, they plan to make genes encoding viruses that can control the fungi responsible for Dutch elm disease and certain crop diseases. At the same time, they hope to use such viruses to learn more about how fungi do their damage. - E. Pennisi