Sequencing the human genome seemed like a discovery so important that it couldn’t be overhyped—we had, after all, transcribed the blueprint for human life—but biotech executives somehow managed the trick. William Haseltine, the founder of Human Genome Sciences, predicted in 2000 that he would halve the time and money required to bring a drug to market. Randy Scott of Incyte Genomics claimed that, “In 10 years, we will understand the molecular basis for most human diseases.”
Not quite. The cost of bringing a drug to market has increased dramatically, quibbles about accounting methods notwithstanding. The process still takes more than a decade. We already had a thorough understanding of diseases linked to single genetic sequences, such as Huntington’s and cystic fibrosis, but if anything, exploring the genome has taught us how complicated the relationship between genes and diseases really is. Last year, for example, researchers in Canada linked 71 genetic regions to inflammatory bowel disease, bringing the total to 163 and counting.
Almost 13 years after the map of the human genome was published to great fanfare in Science and Nature, it’s fair to ask where all the miracle drugs are. The 10 leading causes of death have changed very little since 2000. Life expectancy has risen by 1.9 years, but much of the change is a result of improved health among minorities rather than pharmaceutical breakthroughs.
What happened? Researchers started with an oversimplified view of the genome’s role in human health.
“The genome provided a full parts list for the first time in biology, which was a huge contribution to biology and also drug discovery, but it didn't describe how things fit together or worked together, and that was a big problem,” says University of California–San Francisco pharmacologist Brian Shoichet.
Another excuse is time. Since developing a drug takes at least 10 years, it’s arguably unfair to expect many breakthroughs already. That line of reasoning, however, forecasts an imminent pharmaceutical revolution. I’d suggest that you not hold your breath, in part because researchers spent the first few years after 2000 chasing shadows.
Before the genome was sequenced, researchers estimated that humans had about 100,000 genes, based on the variety of proteins in the body and the now-laughable assumption that complex organisms must have more genes than simpler forms of life. (Grapes have more genes than humans do.) Many pharmacologists believed that 100,000 genes meant 100,000 potential drug targets. Even after we learned that there are only about 21,000 genes, drug companies salivated over the possibilities.
“We saw a brand-new landscape and tried to explore everything at once,” says Andrew Hopkins, who studies computational pharmacology at the University of Dundee, Scotland. “Most of what we saw turned out to be boggy fens and perilous cliffs. We didn’t know where the good real estate would be, and the price of innovation was failure.”
Hopkins is now famous among pharmacologists for developing a concept called the “druggable genome.” In 2002 he argued that only about 10 percent of human genes produce proteins that could bind to a small molecule, which is the form most drugs come in. Of that number, only 20 percent of genes are likely to be associated with disease. In the end, there are probably only a few hundred “druggable” targets, and we’ve already found drugs that work on about half of them. The field narrowed considerably.
Hopkins’ sober assessment came too late for many drug companies, which were investing heavily while attempting to beat each other to new drug discoveries. As the folly became clear, drug executives at Pfizer used to joke that the value of a new genome-derived target was negative $30 million. Drug companies drowned in opportunities. The industry laid off more than 230,000 workers between 2005 and 2010, due in large part to those failures.
The genome map has, undeniably, produced several winners. Cancer drugs, which are often fast-tracked by the FDA, have seen the most movement since the genome was sequenced. Many of the new drugs are biologics, which use living cells or substances derived from living cells to target cancer. The human genome map contributed significantly to this field. The genome has also led to drugs that work on the same principle as pre-millennial versions, but with fewer side effects.
Consider the story of Belviq, the trade name for the drug lorcaserin. Although the FDA approved the drug in 2012, its story really began in the mid-1990s. You may remember the weight loss drug Fen-Phen. It was designed to act on serotonin receptors in the part of the brain that controlled appetite. It did that quite well—hence its enormous commercial success—but it also affected serotonin receptors in the heart. Some people taking the drug developed heart valve problems, and to this day, Pfizer, which owns the company that manufactured Fen-Phen, is dealing with the fallout. Pfizer has been ordered to pay hundreds of millions of dollars in damages. The disaster came down to our misunderstanding of serotonin receptors.
“When I started studying serotonin 30 years ago, we thought there were only two receptors,” said Bryan Roth, a pharmacologist at the University of North Carolina. “I remember going to conferences where people argued over which was the real receptor.”
During the course of mapping the human genome, researchers noticed several genetic sequences that appeared to code for additional serotonin receptors. We now know that there are at least 14 serotonin receptors, and it’s possible for a drug to act on one without acting on the remaining 13. That’s where Belviq came in. Using the information gleamed from the genome, pharmacologists were able to hit the appetite-controlling cells without damaging the heart valve.
Belviq is unlikely to have a major impact on human life expectancy. As Lindsay Beyerstein wrote in Slate when the drug came out, patients experience a 3 to 3.7 percent weight decrease, which is little more than a rounding error in the era of gastric bypass surgery. But the success of Belviq is hugely important to pharmaceutical researchers. It has given them a process to follow, rather than chasing proteins all over the human genome.
“Find out where the receptors are expressed, identify an orally available molecule that acts on it, then prove its clinical efficacy in animal models,” says Dominic Behan, chief science officer and co-founder of Arena Pharmaceuticals, the maker of Belviq. The company has a handful of other drugs in the pipeline based on the same simple technique.
Hopkins, the skeptic who pointed out how few druggable targets were to be found in the genome, says the real revolution will be in finding the right patients, rather than finding the right drugs.
“Sequencing one genome in 2000 allowed us to identify drug targets, but the bigger deal is sequencing everyone’s genome, and doing it cheaply,” he says. Doctors have been trying for years to subdivide patients into risk groups or identify certain kinds of patients who were more likely to benefit from a drug. The problem was that our classification methods, which were based on things like race or gender, were pretty clunky. In theory, sequencing individual patients’ genomes could allow researchers to see exactly why a drug performs better in some groups than others, rather than groping around in the darkness of simple correlation.
Keep in mind, though, that personalized medicine was high on the agenda back in those heady millennial days when the genome was sequenced. It should serve as another reminder that the imaginations of doctors and pharmacologists sometimes run ahead of the science.