When it comes to drug development, Jon Thorson studies sweet spots.

The UW-Madison pharmacy professor works with sugar molecules that enable natural compounds to fight cancer or infections -- or avoid unwanted side effects.

Thorson and his colleagues have developed two ways of achieving a process called glycorandomization. It involves adding or changing the sugars attached to compounds to make the compounds more effective or less toxic.

The hope is to improve on drugs such as vancomycin, an antibiotic, and digitoxin, a heart drug that has also shown promise in treating cancer. New drugs also could be discovered by screening hundreds of new variants of compounds created through manipulation of their sugars.

The work involves carbohydrate chemistry that big pharmaceutical companies largely avoid because it requires many steps, Thorson said.

His techniques provide shortcuts. An enzymatic method he created, for example, relies on bacteria such as E. coli to do the heavy lifting.

"You just feed sugars and your molecule of interest to the bacteria and it will spit back out your processed molecules," Thorson said.

A chemical method, which uses a chemical structure that only reacts with sugars but many different kinds of sugars, does largely the same thing in one step, instead of the five to 10 steps required by more traditional approaches.

In a paper published last September in the journal Nature Chemical Biology, Thorson and his colleagues reported on a new enzyme they created for the sugar work. In a paper in the journal Science in 2006, they described a way to harness the catalytic activity of enzymes to take a sugar from one compound and hand it over to another.

The discoveries themselves are being harnessed by Centrose, a biotech company Thorson started in Madison in June.

On campus, Thorson is active in the National Cooperative Drug Discovery Group, which has a $5.6 million federal grant to develop natural compounds into drugs.

Through the program, he gave digitoxin a makeover, replacing its three sugar molecules with one, which might reduce its side effects if used as a cancer drug. The National Cancer Institute is about to start animal studies on the reconfigured digitoxin.

Thorson works closely with Ben Shen, a pharmacy professor who heads up the drug discovery group.

"We want to seal the gap in drug discovery between academic research and industry," Shen said of the program.

Shen, who came to UW-Madison in 2001, the same year Thorson arrived, said Thorson is "one of the most creative scientists I've ever met."

Had Thorson not spent a summer in college at the National Institutes of Health, he would be prescribing drugs today instead of developing them.

After being accepted into medical school, Thorson shadowed a cardiac surgeon and volunteered for clinical studies at NIH, where he realized he was more intrigued by the research side of medicine.

He went to graduate school in chemistry and came to UW-Madison after a stint at the Sloan-Kettering Institute in New York City.

Many of the drugs people use every day come from natural compounds, Thorson said. He believes there is great potential to improve on them and find more.

"There's a lot of untapped science there," he said.