New Haven, Conn. -- Many medications work by binding tightly to a protein and disrupting its normal function, including those that prevent transplanted organs from being rejected, keep HIV infections in check, and even stimulate nerve regrowth in spinal cord injuries. At Yale University, chemists are putting that fact to good use to help design more effective pharmaceuticals with the help of computers.
"Stronger chemical binding generally leads to a more effective drug that can be given in lower dosages, thus reducing unwanted side-effects," says Yale chemistry professor William L. Jorgensen, who specializes in tweaking chemical compounds to tighten their ties to target proteins. Using sophisticated computer graphics and his knowledge of the three-dimensional structure of protein molecules, Jorgensen consults with major pharmaceutical firms regarding new methods of rational drug design.
"Historically, drug development has been largely trial and error. To improve a drug meant making a very large number of modified compounds and laboriously testing each one," Jorgensen said. "Part of the hope with rational drug design and computers is to narrow the search, to pinpoint the modifications that will work best."
The Yale chemist begins with a three-dimensional picture of a drug bound to its protein target -- usually generated through a process called X-ray crystallography -- then modifies the compound and computes how the changes affect the drug's binding ability. The 3-D structures of hundreds of key human proteins have been solved in recent years, many of them by Yale scientists, making rational drug design possible.
Jorgensen also has developed widely used computer programs, such as BOSS and CAMEO, that enable scientists to model complex organic solutions and predict the products of chemical reactions, if given the starting materials and conditions. In recognition of his contributions in this field, he will receive the 1998 Award for Computers in Chemical and Pharmaceutical Research on March 31 at the annual meeting of the American Chemical Society (ACS) in Dallas. The prize is sponsored by IBM North America, Scientific & Technical Systems & Solutions.
The diseases that may eventually be treated more effectively because of Jorgensen's computer programs range from dementia to diabetes. Scientists recently discovered, for example, that some medications that bind to a protein called FKBP can stimulate the growth of nerve and brain cells, which makes them excellent candidates for treating spinal cord injuries as well as Parkinson's and Alzheimer's diseases. FKBP also plays a role in transplant rejection, making it a doubly important protein to control.
Jorgensen and his co-workers have performed calculations on FKBP with a series of potential drugs, analyzing why they vary in binding strengths and suggesting ways to improve their design -- work that Jorgensen will describe on March 31 at the ACS annual meeting. The Yale chemists also have analyzed drugs that regulate the protein thrombin, which is important for controlling blood clotting.
Another application being explored in Jorgensen's laboratory is the development of improved anti-HIV drugs that counter mutant strains that are resistant to current drugs, such as AZT and the protease inhibitors. Jorgensen's calculations also have been applied to potential drugs that can regulate the natural synthesis of insulin for diabetics in a gene therapy approach involving implanted cells.
Vast Increases in Computer Speed
Vast increases in computer speed in recent years make it possible to perform complex computations of chemical processes in minutes instead of days, said Jorgensen, who has created a supercomputer in his Yale laboratory by linking together 25 Pentium machines in a network. He and his colleagues now can analyze between 50 and 100 chemical variations, or analogs, of a leading drug each week, he said. Drug companies frequently generate as many as 1,000 analogs of a leading drug, such as cyclosporin used to prevent transplant rejection, in an effort to find a more effective compound.
Jorgensen said his work also deals with broader chemical problems, such as calculating how changing a solvent in an industrial process -- from water to methanol, for example -- will change the rate and products of reactions. Using powerful computers, he can model liquid water by treating mathematically the interactions between thousands of individual water molecules, thus computing the density of water at any temperature or pressure.
"When we began our work 20 years ago, computer simulations had only been carried out for some simple liquids. With the development of the BOSS program, which performs Monte Carlo statistical mechanics calculations, we were able to model numerous complex organic liquids and solutions. These studies provided great detail in the structure of liquids and evolved into our drug design work," said Jorgensen, who recently was ranked among the top 20 chemists in the world in the number of times his work was cited by other scientists. The ranking is calculated by the Institute for Scientific Information.
###
Note to Editors: William L. Jorgensen, the Whitehead Professor of Chemistry at Yale University, received his Ph.D. from Harvard University, where he worked with Professor E. J. Corey in the use of computers to design synthetic routes for making organic molecules. (Corey later received the 1990 Nobel Prize in Chemistry.) While at Harvard, Jorgensen developed the first computer-generated 3-D renderings of molecular orbitals. His research at Yale is supported by the National Institutes of Health and the National Science Foundation. For interviews, contact him at (203) 432-6278.
A color slide of an FKBP protein molecule binding with a drug that stimulates nerve growth is available from the Yale Office of Public Affairs upon request. The computer-generated image was produced by X-ray crystallography. Call (203) 432-1326.