Science / Medicine : PROBING...
Scientists using new computer technologies are gaining insight into the three-dimensional structure of some of the body’s key proteins. At the forefront of this new frontier is one of these proteins--the enzyme superoxide dismutase (SOD), famous for preventing oxygen toxicity in the body and thought to hamper such diverse disorders as heart disease, cancer, Alzheimer’s disease and the aging process.
The scientists have already uncovered a complex scenario in which the enzyme lures and traps the superoxide molecule, which it is programmed to destroy. By gathering similar information about a number of crucial proteins, the researchers hope to create “designer proteins” that could fight cancer and AIDS, among other diseases.
Although the genes behind many types of cancers and other diseases have been uncovered in recent years, without knowing the structure and function of the proteins these genes commandeer cells to generate, scientists are unable to concoct compounds that combat the root cause of the disorders.
Grasping the three-dimensional structure of a protein is a daunting task. Most proteins house thousands of atoms that link together to form chains. These chains wind in and out of each other, forming a structure that looks like a confused jumble of string.
To add to the complexity, a protein changes its shape several times within a fraction of a second, and a change in a single amino acid (a small grouping of atoms), can dramatically alter the protein’s ability to do a specific task, like latch on to a virus.
For decades biochemists have bombarded proteins with X-rays to get a picture of their three-dimensional structure. Such work led to the discovery of DNA’s form and function, which snowballed into genetic engineering. But such X-ray diffraction just “gives you a lot of numbers that tell you where the atoms are, but not what they are doing,” says Scripps Clinic biochemist Elizabeth Getzoff.
To help with that task, biochemists are turning to computers to provide views of the unseen world of protein dynamics. A new technique that will essentially take moving pictures of proteins as they do their various biochemical chores also holds promise for animating this miniature yet vital world.
Innovative computer graphics were critical to Getzoff and her colleague John Tainer’s unraveling of how SOD lures superoxide molecules close and then nabs them. In a dazzling array of dots, a computer using X-ray diffraction data displayed the molecular surface of SOD. Each dot was color coded according to charge.
Negatively charged regions of the enzyme were red or yellow whereas positively charged sections sported blue or green colors. This information had been available in numerical form before, but without the graphic description made possible by the computer, the scientists could not see an emerging picture of the workings of the protein.
What leaped out of this graphic was a large patch of blue in the center of the molecule. This blue cloud of dots lined a channel formed by two loops of amino acids. At the bottom of this channel a positively charged copper ion waits to merge with the oppositely charged superoxide and start the chemical chain of events that splits the toxic superoxide into two harmless compounds.
To the display of the crucial channel, the scientists added myriad vectors, or arrows, which pointed toward the positively charged regions of the channel and away from the negatively charged sections. Since superoxide is attracted to positive charges and repelled by like negative charges, the arrows essentially pointed out the path the unsuspecting molecule would take as it approached the enzyme.
“The direction of vector flow indicated that an oxygen radical would be repelled from the banks of the channel,” Getzoff says, “and drawn into the channel itself, where it would be swept along before diving down to dock next to the positively charged copper ion.”
Certain amino acids in SOD play key roles in steering the oxygen radical to the copper ion at the bottom of the channel, the researchers discovered. When just one of these was deleted from the arrow model, superoxide took a totally different path and never met up with the copper ion.
Revealing how exactly the oxygen radical binds to SOD, a computer graphic displayed the molecular surface of the channel. This image showed that nestled in the narrowest part of the channel floor where the copper ion lay in wait were two distinct potholes. Using a joy stick, the researchers tried to dock a surface model of superoxide into one of the potholes. To their delight, it fit perfectly.
“It was really exciting to discover this elaborate trap that SOD sets for superoxide,” Getzoff says. The California scientists made a computer graphics video of this dramatic molecular encounter, which they appropriately called “Terms of Entrapment.” The video is part of a computer graphics show that is currently touring various museums in the country.
But although computer models, such as those used by Getzoff, may actually predict what a protein is doing, no one has yet been able to see a protein in action, at least in three dimensions. That situation may change now that biochemist Keith Moffat and his colleagues at Cornell University have developed a system that should allow them to take “movies” of antibodies binding to viruses, enzymes devouring compounds, and other critical reactions that occur extremely rapidly in the body.
The system, which combines a particle accelerator with a device called an undulator, generates large and highly focused quantities of X-rays in frequent bursts, allowing millions of consecutive pictures of a protein to be taken within just half a second. The researchers expect this system will capture a protein in motion as it performs some vital role.
“This is an important goal,” Moffat says, “because the essence of life is movement, whether at the molecular level or at the level of the whole organism.” The system should be particularly useful to drug companies because it will be able to pinpoint the intermediate three-dimensional structures a protein takes as it performs some biochemical chore. Scientists currently cannot see many of these intermediates because they last for such a tiny fraction of a second. But drug developers are eager to get their hands on them because a drug that mimics an intermediate could be used to block an enzyme’s action, for example, or to stop a virus from infecting cells.
Before Moffat can start his molecular movie business, however, he needs to concoct some chemical systems that allow reactions to take place within the crystals used to take the X-ray pictures. He has already shown that it’s feasible to take X-ray diffraction pictures on the short time scale needed to animate a chemical reaction and he expects to have the whole system up and running within a few years.
In the meantime, drug companies are gearing up to use the three-dimensional structures of key proteins as starting points for the design of new drugs. This approach runs contrary to the standard hit or miss technique of haphazardly screening several compounds to see if they can do something useful in the body. According to biochemist David Matthews of Agouron Pharmaceuticals in San Diego, the new approach “is going to revolutionize the way drugs are developed, because its only limit is the imagination of the organic chemist.”
Using the three-dimensional structure of an enzyme critical to the spread of tumor cells, Matthews designed several compounds that show promise in being more effective and rendering fewer side effects than most current anti-cancer drugs. “These compounds are completely novel structures,” he says. “Nobody in their right mind would have ever predicted these molecules could bind to the enzyme had they not known its structure.”
The compounds, still in the experimental stage, were designed to “fit into the nooks and crannies of the enzyme” blocking it’s ability to bind to certain substances. Matthews also added just the right type of side chains to the compounds so they could easily slip into the solid tumors typical of head, neck and colon cancers. Most anti-tumor drugs can’t penetrate these tissues.
Given the promise of an understanding of a protein’s structure, the National Institute of General Medical Sciences (NIGMS) recently earmarked funds for studies aimed at uncovering the structure of proteins generated by HIV, the virus that causes AIDS. A structure of one of these viral proteins may surface by the end of the year, according to Marvin Cassman, director of the biophysics and physiological sciences program of NIGMS. The next step will be to devise drugs that block the protein’s ability to foster HIV’s infection of human cells. “We’re working at the edge of what is currently possible,” Cassman says.
