Drugs By Design

Thanks to genetics, the pharmaceutical industry is exploding with new ideas

By Christine Gorman Monday, Jan. 11, 1999

It is the year 2025, and some things haven’t changed. The sky is still blue. The Dow is poised to set another record. And Jose Rodriguez (Michigan State, class of ’04) has just learned that he has colon cancer. But he’s not too concerned. Thanks to the genetic revolution that swept over the pharmaceutical industry 30 years earlier, scientists have developed a variety of anticancer drugs that work far better, and with fewer side effects, than the old poison-and-burn treatments of the late 20th century.

The oncologist takes a few cells from Jose’s tumor and places them on a microchip. Within minutes, the chip identifies five mutant genes that, like some kind of diabolical cheerleading squad, have pushed Jose’s cancer to grow, grow, grow. Someday, perhaps soon, doctors will be able to fix the wayward genes themselves. Until then, they will have to rely on the next best thing: drugs developed by pharmaceutical firms that block the destructive messages generated by the errant genes. Jose’s physician selects a combination of treatments that matches the tumor’s genetic profile. Six months later, no trace of Jose’s cancerous growth can be found.

That scenario is not as farfetched as it sounds. Talk to anyone in the pharmaceutical industry, and you’ll soon discover that genetics is the biggest thing to hit drug research since a penicillium mold floated into Alexander Fleming’s petri dish. Sure, scientists have long known genes play a role in almost every ailment from Alzheimer’s to yellow fever. But it is only in the past few years that they’ve learned how to use that information to identify a multitude of new targets and pathways for drug design. Let’s count the ways.


Geneticists estimate that there are 2,000 to 5,000 genes that either cause, or predispose humans to, various diseases. In practical terms, that means there will be many, many more potential avenues of research than the entire pharmaceutical industry could possibly hope to investigate over the next 20 years. Each company has a different strategy for exploiting that bonanza, and most are more than happy to tell you what’s wrong with the other guy’s approach. But they all agree on a few key points:

–Drugs will be safer, more powerful and much more selective than ever before.

–Doctors will be able to consult your genetic profile to determine ahead of time whether you are more likely to respond to one type of medication or another.

–Computers and other digital technologies are going to play a much bigger role in evaluating new research and determining how patients should be treated.


To understand how this promising future might come to pass, it pays to review a little history. Back in the old days–which is to say just a few decades ago–the process of discovering a new drug was a lot like shooting a quiver of arrows into the air and then running around to see what they hit. Occasionally scientists would get lucky, as Fleming did in 1928, but most of their efforts were wasted.


Drugs By Design

Thanks to genetics, the pharmaceutical industry is exploding with new ideas

By Christine Gorman Monday, Jan. 11, 1999

(2 of 5)

The odds started to improve in the 1970s and early ’80s as researchers used recombinant-DNA technology to mix and match bits and pieces of hereditary material. Suddenly they had a front-row seat from which to watch genes direct the construction of RNA molecules, which in turn assembled proteins, enzymes and other biological molecules. Instead of shooting their research arrows into the air, drug companies could take aim at defined targets. Focusing on serotonin receptors in the brain, for example, led to the development of Prozac and its chemical cousins for the treatment of depression. Targeting histamine receptors in the stomach produced Tagamet and then Zantac to relieve acid indigestion.

By the 1990s, decades of work had led to the identification of 500 different biological targets for drugs. Thanks to the Human Genome Project, researchers expect to identify another 500 in just the next few years. Soon there will be more new targets than even the largest companies can handle. Then the trick will be to figure out which targets to go after first, and how.

One approach is to focus on the diseases that affect the most people–those associated with aging, say–and to do it by aiming for the targets that are the most accessible. That generally means designing a drug that affects the proteins and enzymes that sit on a cell’s surface or in its cytoplasm, not the genes that code for those proteins and enzymes, which are usually tucked away in the protective nucleus of the cell. This is the strategy favored by such big, traditional drug companies as Merck, Pfizer and Novartis–though it is by no means the only game in town.


While the pharmaceutical giants are eager to exploit the latest genetic information to create new drugs, they don’t see the need to reinvent the wheel completely. The medications they design will still be derived from chemical compounds, or “small molecules” in industry parlance, that happen to be biologically active. (In fact, most of the drugs developed over the past 100 years, from aspirin to Zoloft, are small molecules.) Among the advantages: small molecules aren’t destroyed in the stomach, so they can be taken by mouth. Furthermore, they don’t get noticed–or attacked–by the immune system. Two of the most active areas of small-molecule research are Alzheimer’s disease and cancer.

In 1992 Dr. Allen Roses, then at Duke University and now at Glaxo Wellcome, discovered a link between a particular protein in the blood and the risk of developing Alzheimer’s disease. The protein, called apolipoprotein E, works like a cargo ship ferrying cholesterol around the body–a task that seems, at first glance, to have little to do with a degenerative condition in the brain. But 67% of Alzheimer’s patients carry a gene that codes for one version of the protein, called apo E4, in contrast to 30% of healthy adults. So although most people with apo E4 never develop Alzheimer’s, a significant fraction of them will.

Roses believes he won’t have to figure out exactly why apo E4 increases the chances of developing Alzheimer’s. As long as he can determine how the brain uses it differently from other versions of the protein, he should be able to develop a drug that either enhances or reduces that effect. The new drug may not be able to treat everyone with Alzheimer’s, but at least it could help some.

Drugs By Design

Thanks to genetics, the pharmaceutical industry is exploding with new ideas

By Christine Gorman Monday, Jan. 11, 1999

(3 of 5)

Finding a likely target, of course, doesn’t guarantee success. Consider colon cancer: scientists believe at least three things have to go wrong for colon cancers to form. They liken the situation to a car accident. One of the genes that tells cells to divide (the accelerator) must get stuck in the “on” position. Another gene that tells cells to slow down (the brake) must be disabled. And the molecules that fix any mistakes in the DNA code (the repair crew) have to go on strike. In half of all colon cancers, the accelerator is a gene called ras, which makes a protein that stimulates cell growth. It was the ideal target for an anticancer drug.

Or so it seemed. “We banged our heads against the wall for 10 years,” says Dr. Alan Oliff, head of cancer research at Merck. “We were on the verge of abandoning the project.” Then Oliff’s team realized something critical: the ras protein can’t do its job until it has been activated by another enzyme called a farnesyl transferase. Maybe that would make a better target? Early word is that it does, but Merck won’t publish the findings from its first human trials until sometime next year.


Whereas traditional drug companies focus on developing chemical compounds, the biotech industry prefers to use biological ones–hormones, proteins and other substances that either already exist in the body or can be created from scratch. Examples include interferon, the clot buster tPA and the new breast-cancer drug Herceptin.

But even among the rarefied biotech elite, there are mavericks who think they have a better idea. They want to move one step closer to the gene by targeting the RNA molecules that transfer information from genes to proteins. And they have the perfect molecular tool with which to do it. By synthesizing strands of DNA that are the mirror image of the RNA they wish to block, researchers can produce a drug that is more specific than anything else on the market. Because it interrupts the “sense” that the cell is trying to make of the RNA molecule, the new technology is called, appropriately enough, anti-sense.

There are still some kinks to work out. For one thing, the body’s own immune system often attacks the anti-sense DNA, mistaking it as a potentially harmful virus. For another, many cells in the body don’t allow the anti-sense molecules to cross their membranes. “Nine years ago, everyone thought, wow, this is dynamite,” says Dr. Art Krieg, editor of the journal Anti-Sense and Nucleic Acid Drug Development. “Then they ran into technical hurdles, and the pendulum swung the other way.” Now, says Krieg, a few anti-sense compounds are starting to show promise. Among them is a drug called Vitravene, which was approved by the Food and Drug Administration in August and is used to prevent blindness in AIDS patients infected with cytomegalovirus.


Genes don’t just tell you how to make drugs. They can also tell you whom to treat.

Drugs By Design

Thanks to genetics, the pharmaceutical industry is exploding with new ideas

By Christine Gorman Monday, Jan. 11, 1999

(4 of 5)

All drugs have some side effects. By scanning a patient’s genetic profile, drug companies may soon be able to figure out ahead of time who is most likely to suffer an adverse reaction. Case in point: Abbott Laboratories has an experimental treatment for asthma that triggers liver abnormalities in about 3% of patients. But it seems to work pretty well in everyone else. So Abbott has asked the French company Genset to see if it can develop a genetic profile of those patients who should never take the medication. The technology isn’t foolproof, but it may give Abbott the tools with which to market its drug more safely.

Knowing what’s in your genes could also take some of the routine guesswork out of medicine. If you’re diagnosed with high blood pressure, for example, your doctor may have to try three or four different pills before finding one that works for you. That’s because blood pressure is controlled by probably dozens of different genes, any one (or more) of which may be responsible for your particular condition. By screening your DNA and comparing your genetic profile to those of patients who have already responded to particular medications, your doctor may be able to prescribe the right drug the first time around. The money you save would come at the expense of the drug companies, of course, since they would no longer profit from any trial-run prescriptions.


Focusing on one or two genes and the proteins they code for has already started paying off in the search for new medicines. But the future of drug discovery is going to be centered on a better understanding of complex biological networks like the brain and the immune system. “The only way you can understand complex systems is to look at many genes and proteins at a time,” says Lee Hood, chairman of molecular biotechnology at the University of Washington in Seattle. How many? Perhaps 1,000, or 10,000, or even 100,000.

Enter the microchip. Just as chips made of silicon allow computers to process millions of bits of information at a time, chips that process or even incorporate fragments of DNA will one day analyze millions of genetic sequences simultaneously. Patterns that would otherwise take decades to discern could show up in minutes on a gene chip. Doctors will use gene chips to screen their patients for thousands of genetic defects at once. Pharmaceutical researchers will use them to identify which genes are turned on or off in any given disease or system of the body and therefore might make good targets for drug development.

At least that’s the theory. Gene chips are so far out on the cutting edge that even many scientists have a hard time believing they’ll work. Steve Fodor, CEO of Affymetrix, is used to addressing such doubts. His company, based in Santa Clara, Calif., is widely regarded as the leader in developing gene chips. “We’ve had to define a lot of new technology, terminology and applications,” he says. “But a fantastic new field has sprung up.”

Drugs By Design

Thanks to genetics, the pharmaceutical industry is exploding with new ideas

By Christine Gorman Monday, Jan. 11, 1999

(5 of 5)

So how would you make a gene chip? Let’s say you want to identify which genes get turned on, or “expressed,” by the immune system in the first few weeks after the AIDS virus begins its attack on the body. First you download the sequences of perhaps 10,000 genes–every A, C, G and T of the hereditary alphabet–into a computer. Then, still using the computer, you figure out what the mirror image of each sequence would be. (DNA can mirror itself as well as RNA.) The aim is to transform the mirror-sequence data into actual strands of DNA that are planted like rows of corn on the glass bed of a chip. Each strand is built up, letter by letter, in much the same way the layers in a silicon chip are created.

Once the strands are complete, the gene chip is ready for use. You take a sample of blood from a patient who has just developed a raging HIV infection. Various genes in his immune system are churning out millions of RNA molecules that will assemble the proteins needed to combat the infection. You extract the RNA and break it into pieces, tag each piece with a fluorescent chemical and pour the whole mess over the gene chip. The RNA tightly binds only to its exact DNA complement on the chip. The fluorescent tag tells you where on the chip you have a match. Then you look up the sequence of each matched spot on the chip and read out a precise catalog of which genes are being expressed. By comparing the results from several patients–some of whom are more successful at fighting the virus than others–you may be able to identify targets that could lead to powerful new anti-AIDS drugs.

Such feats of computational biology are still a few years off or, in the worst case, maybe even a few decades away. The point is, we are just beginning to see how dramatically gene-based science can change the ways in which new drugs are discovered and developed. Blind luck will play an increasingly smaller role as scientists tease out the complex interplay between genes, proteins and the environment. There is going to be confusion–some setbacks and disappointments–at least at first. But most in the field agree that pharmaceutical research has finally entered its golden age.

–With reporting by Dan Cray/Los Angeles, Bruce Crumley/Paris and Alice Park/New York