Learning Bacteria’s Secrets
While doctors and public health officials were becoming aware of the extent of the drug-resistance problem, scientists and drug companies were taking note of the growing stockpile of knowledge pouring in from gene-sequencing projects. J. Craig Venter, chief executive officer of The Institute for Genomic Research in Rockville, Md., estimates that the biochemical codes for up to 40,000 new microbial genes already exist in public databases and that researchers will discover as many as 500,000 genes in the next decade, mostly in microbes. Now, says Venter, industry must winnow its interest to the few that will make the best drug targets.
Of particular interest to the pharmaceutical industry are the fully sequenced genomes already available for several bacteria. Understanding which genes, and therefore which proteins, make a bacterium function is essential to learning how to fight it. Having complete genomes “permits entirely new forms of analysis,” says David Searls, a vice president at SmithKline Beecham Pharmaceuticals. “With a whole genome at your disposal, you have in a sense a closed world. You can analyze an organism in terms of biochemical pathways, for example, and be assured that you’re going to have represented in your database every possible biochemical reaction.” And that is why gene sequencing became the second motivator for drug makers to develop new antibiotics and rescue those in distress on the market.
Last summer at a forum on resistant bacteria held by the National Academy of Science’s Institute of Medicine, gene sequencing pioneer Venter described data released from four of the first completely sequenced genomes of bacteria. Then he told his colleagues: “The chief breakthrough that allowed this to take place was not molecular biology. It was actually bioinformatics that allowed us to deal with thousands and thousands of sequences.”
In August 1987 no one had heard of bioinformatics. Ten years later an Internet search engine came up with more than 22,000 references to it. This field of study was born when scientists began to realize that the volumes of data to come from gene sequencing would be of little use unless a systematic process could be devised to organize and analyze them. Using computer programs, researchers can compare sequences of newly discovered genes with known ones from other species. To experts, a close match may offer clues to the functions of the “new” genes. Genes that code for proteins are the target of much gene sequencing. Scientists can also use bioinformatics to determine “where and when the specific messages coding for these proteins are made,” explains George S. Michaels, one of the early bioinformatics specialists.
And so this tool quickly became the third factor prompting the drug industry’s renewed research interest in antibiotics. Almost overnight, bioinformatics is becoming a profession. Michaels, who now heads George Mason University’s graduate program in bioinformatics, says his four-year-old program awarded its first Ph.D. last year. And in recognition of the importance the technology plays at SmithKline, last year that company named David Searls vice president for bioinformatics. “I think that genomic sequencing and bioinformatics opened up really unparalleled opportunities in antimicrobial research,” says George H. Miller, presidential fellow-vice president for microbiology at the Schering-Plough Research Institute. “You can select new targets now.”
Given the resistance problem and the new tools to try to do something about it, “there’s a lot more enthusiasm” to create new antibiotics, says Allan Weinstein, vice president for international medical and regulatory affairs at Eli Lilly and Company. “You can look at current antibiotic therapy as sort of a sledgehammer,” he explains, while the new approach “directed at specific genes necessary for bacterial life would be more of a stiletto.”
The fourth factor behind the enthusiasm to search for new antibacterial drugs involves two other new tools that have been developed in the last 5 to 10 years that serve to speed up the hunt. In the past, chemists randomly prepared compounds one at a time to see if they had any usefulness as a drug. Now, explains Ted McDonald, director of chemistry at Pharmacopia, Inc., in Princeton, N.J., most drug companies have replaced that method with a much faster one. Through a process called combinatorial chemistry, pharmaceutical chemists create collections of small organic compounds in which they systematically vary the different units or building blocks in the compound’s molecular structure, preparing many possible combinations.
Another technology called high-throughput screening allows these synthetic molecules to be screened in large groups. In one method, plates of 96 miniature wells containing a molecule and a protein target from a bacterium are passed over a detector that shows a color change when a molecule binds to a protein. Some companies prefer attaching molecules to polystyrene beads because of the ease of purifying the product after several chemical reactions by simply washing reagents and byproducts off the beads. Whatever method is employed, the bottom line for drug makers is speed. In the time it used to take to screen a few hundred compounds for potential drug activity, pharmaceutical companies now can screen tens of thousands.