Contact

Duke Health News

919-660-1306

DURHAM, N.C. -- For the first time, biochemists have used computational design methods to transform a protein devoid of catalytic activity into a biologically active enzyme. When the Duke University Medical Center biochemists inserted their artificial enzyme into a microbial cell, it functioned to maintain the cell as would a natural enzyme, even though its structure was much different.

The researchers said their achievement represents an important milestone in the development of their protein design technology, which will lead to an era of "synthetic biology." Such capabilities, they said, will enable design of proteins and even whole microorganisms that could find wide use in medical, industrial and environmental applications. Such uses might include living drug or chemical factories, or microbial "sentinels" that could sense pollutants or toxic chemicals.

The researchers, led by professor of biochemistry Homme Hellinga, Ph.D., published their findings in the June 25, 2004, issue of the journal Science. Besides Hellinga, other co-authors are Mary Dwyer and Loren Looger. Looger is now at the Carnegie Institution department of plant biology in Palo Alto. Their research was sponsored by the Defense Advanced Research Projects Agency.

Enzymes are the protein workhorses of the cell, catalyzing the myriad of chemical reactions that support life. They are synthesized as long strings of amino acids, which fold origami-like into functioning globular shapes. While some proteins form structural components of the cells, others act as enzymes -- attaching to specific target molecules called substrates and using chemical energy to transform those substrates in some way. Such attachment and chemical catalysis takes place in a pocket of the molecule called its active site.

Hellinga and his colleagues used their computational design methods to transform a non-catalytic protein called a ribose-binding protein into an enzyme known as triose phosphate isomerase. Isomerases act on molecules to rearrange their geometric structure -- like converting a left-hand glove into a right-hand glove.

"The conversion of a protein into a catalyst presented the most challenging design problem we have yet faced," said Hellinga. "Our previous work has concentrated on creating proteins with highly specific binding of molecules drastically different from their normal targets. However, with catalysis, we faced not only the challenge of binding specificity, but of having the protein then perform chemistry on the substrate molecule and of stabilizing the fleeting 'transition state' between the original molecule and the final product."

Beginning with the ribose binding protein, the biochemists first used their computational methods to redesign the protein's active site so that it would bind both the initial substrate and the product molecule.

Their computational method involves basically "mutating" a protein inside a computer by selectively altering its individual amino acids to work toward a protein whose active site has the three-dimensional shape and chemical binding properties that fit a specific target molecule. The method not only generates the design of an array of candidate mutated proteins -- which can be vast -- but narrows down that array to a manageable number that can be synthesized and tested in the laboratory. The researchers use an array of interconnected computers roughly the equivalent of high-end desktop computers to perform their calculations.

In transforming the binding protein into an enzyme, the researchers arrived at designs involving alterations in 18 to 22 of the 271 amino acids of the ribose-binding protein.

"Once we had a selection of candidate proteins that had the appropriate binding properties, we used a newly developed technique to introduce catalytically active amino acids into the design and screened the resulting compounds for activity," said Hellinga. "Of 14 final designs, we found seven that were quite active enzymatically, and one that enhanced the chemical reaction over a hundred thousand times more than the uncatalyzed reaction."

However, noted Hellinga, the artificial enzyme still remains a thousand times less active than the natural enzyme.

To demonstrate that the artificial enzyme was nevertheless biologically active, the researchers introduced it into a version of the gut bacterium E. coli that lacked the natural enzyme. The resulting bacterium, when exposed to conditions that demanded the functional activity of the isomerase, showed normal growth.

"The real proof of success is whether the artificial enzyme can take over the function of the wild-type enzyme," said Hellinga. "And that's what we demonstrated."

Hellinga and his colleagues are now seeking to apply their enzyme design techniques to other proteins and other catalytic processes.

"The real challenge for the field is to use these techniques to do what's called 'green chemistry,'" said Hellinga. "It's a dream that has been around for a long time -- to have the ability to use engineered microorganisms to synthesize desired chemicals without resorting to large-scale industrial processes that are inherently wasteful and inefficient. We believe such artificial enzymes are an important step toward that goal," he said.

Such green synthesis could enable far more efficient synthesis of products ranging from complex drugs to plastics and fuels, said Hellinga. Current industrial methods produce unwanted side reactions and waste chemicals that must be disposed of.