Through genetic engineering, goldenrod weed and lowly yeast could yield premium natural rubber, lessening America's dependence on imports and petroleum-based synthetics, a U.S. Department of Agriculture scientist says.

Plant physiologist Katrina Cornish is among the researchers looking into ways of getting high-grade rubber from fast-growing plants or from yeast or bacteria in vats. At the Agricultural Research Center in Albany, Calif., she is looking for genes in the tropical rubber tree and in guayule, a desert shrub. In these plants, a gene cues cells to make rubber transferase, the enzyme that forms rubber molecules. Other genes make the plants' rubber production start, continue and stop.Cornish, addressing the First International Conference on New Industrial Crops and Products, in Riverside, Calif., said she wants to insert those genes from the rubber tree or guayule into goldenrod or microorganisms, hoping to make them produce more and higher-quality rubber.

Rubber tree latex, long the primary source of natural rubber, is hand tapped from Hevea brasiliensis trees in plantations, mostly in Malaysia and Indonesia. The United States lacks the climate these plants need.

Synthetic rubber isn't the answer, either, Cornish says. "Natural rubber is more elastic, more resilient and more resistant to heat buildup than synthetic rubber. Also, natural rubber is a renewable resource, unlike synthetics made from petroleum."

Political changes in the rubber-producing countries, crop failures and other possible emergencies make research into the biotechnological production of rubber a good idea, for defense and economic reasons, she says.

The United States imports some 800,000 tons of natural rubber, worth about $500 million, each year for airplane tires, latex surgical gloves and dozens of other products for commercial and defense use.