How Computers Are Helping Drug Companies Find New Medications - WSJ.com
After decades spent searching for compounds in nature that might have medicinal potential, most major drug companies are now also designing medicines virtually. Aided by powerful computational analyses that help identify a target for a drug, researchers manipulate molecules on their computer screens to create custom-made compounds that attack disease-causing proteins.
This method of tailoring molecules is known as structure-based design. It was used to create Xalkori, a therapy launched by Pfizer Inc. PFE +0.50% last year to treat a rare and intractable form of lung cancer. It also has played a role in an Alzheimer’s disease treatment being developed at Eli Lilly & Co, LLY +1.05% an antibiotic crafted by GlaxoSmithKline GSK +2.00% PLC that is in midstage trials and a Sanofi SA SNY +0.61% blood thinner in the final stages of development.
With computer-aided design, pharmaceutical companies are finding treatments they might never have discovered otherwise, even with considerable investments of time and expense, says Magid Abou-Gharbia, the director of Temple University’s Moulder Center for Drug Discovery Research and the former head of drug discovery at Wyeth.
Pfizer
FINDING A FIT Jean Cui (center), a Pfizer researcher, designed the recently approved lung-cancer drug Xalkori.
“You will actually increase the chance of success for coming up with a clinical [drug] candidate” by using it, he says.
Keys and Locks
Structure-based design is quite different from traditional drug-discovery techniques, says Jean Cui, the scientist credited with discovering Xalkori.
Typically, a drug works by attacking a disease-causing protein that is interacting with other molecules in an unhealthy way. By connecting to the protein, a drug can stop it, thereby restoring healthy interactions or compensating for unhealthy ones.
A drug connects to a protein much as a key fits into a lock. For most of their histories, drug makers looked for the keys while ignoring the locks. Drug companies sifted through natural substances found in soil, as well as collections of dyes and industrial chemicals and failed compounds from previous drug research-and-development programs. They would test those samples for any impact on diseased cells and, if they were lucky, find one that worked. This is how penicillin was discovered.
Pfizer
TIGHT SQUEEZE Xalkori works by locking into and blocking a protein essential to a form of lung cancer.
Over the years, companies speeded up the process. They accumulated vast libraries of potential compounds—the end of the Cold War opened up a trove of new molecules collected by Eastern European laboratories—then used robots to quickly run through hundreds of thousands of samples to see if any showed medicinal potential.
Yet, drug screening remained dependent on a company’s good fortune to chance upon a promising compound. When researchers hit on a new medicine, they often wouldn’t know for years why it worked, only that it did. They didn’t know what the key or lock looked like.
Frustrated, a Merck & Co MRK +0.58% . scientist named Joshua Boger left to start a company that aimed to take much of the luck out of drug discovery. His firm would figure what a lock looked like, so it could fashion a better key to fit into it. The company, Vertex Pharmaceuticals Inc., VRTX -2.21% established in 1989, was among several conceived with this mission.
“I felt every project wasn’t using all the information available to it, and one of the kinds of information left out most often was the structural information, telling you what a drug was supposed to do,” says Dr. Boger, now a board member at Vertex, based in Cambridge, Mass., whose hepatitis C therapy, Incivek, was approved last year.
Determining the shape of a lock isn’t easy, however. Because proteins run from X-rays taking their snapshots, scientists must first crystallize them so they can’t escape and then deduce their shape by looking at the patterns left by the X-rays deflecting around them. This requires thousands of interference patterns and powerful computers to analyze them.
Then, researchers must fashion a custom molecule to fit into that lock. Coming up with the right shape can be difficult. Moreover, the molecule also must connect, or bind, to the target. And to be a successful drug, a molecule must have other properties. It can’t be metabolized by the body too quickly or slowly, and it must be able to be synthesized and manufactured in large quantities.
Because of those challenges, structure-based design is more an instrument for boosting companies’ drug-discovery efforts than for revolutionizing them, researchers say. It provides information that is helpful and sometimes crucial for discovering new drugs, but it can’t generate the ideal candidate without other information and work.
“It doesn’t tell us everything, but it gives us a very good clue,” says Tony Wood, who heads chemistry at Pfizer.
Scientist’s ‘Aha’ Moment
Xalkori wouldn’t have been discovered in the early 2000s if not for structure-based design, according to researchers at Pfizer.
In 2003, Pfizer bought Pharmacia, the owner of the biotech firm Sugen, where much of Dr. Cui’s early work on Xalkori took place. Researchers were trying to block a protein called c-Met that was found to play an important role in the growth of cancer tumors. They hit upon a naturally occurring molecule that connected to c-Met, and synthesized some prototype molecules.
But even with further refinement, these prototype molecules didn’t have the properties—such as avoiding quick metabolism in the body—needed to make it a workable drug, Dr. Cui says.
Scientists turned to structure-based design for help. The researchers crystallized the c-Met protein with one of the prototype molecules hooked up to it, fired X-rays at the arrangement and, using computers, deduced the structure of the protein and how the prototype molecule fit into it like a key in a lock.
They emailed the results to Dr. Cui, who began trying to come up with an entirely new molecule that would bind to c-Met and possess properties suitable for a drug. It was difficult, Dr. Cui recalls. The new molecule would have to connect to a site on the c-Met protein that scientists hadn’t expected, and it required a tight squeeze into a small space. Dr. Cui says she found herself trying to solve the puzzle all of the time.
In May 2002, after five months of thinking about it, the design came to Dr. Cui while she was at home watching her two daughters play. The next morning, she took a rough sketch of the design to her boss, and soon colleagues were making compounds virtually on a computer and in test tubes for further study. Within weeks, Sugen decided it would try to turn this molecule into a drug. By February 2003, testing in animals showed that the molecule could stop tumor growth. After the Pfizer acquisition, researchers there further refined, synthesized and studied the molecule until Xalkori emerged ready for testing in humans.
The work showed that Xalkori, known chemically as crizotinib, bound to and blocked the tumor-causing activity of a protein called ALK, as well as c-Met. Xalkori was approved last year for non-small-cell lung cancer caused by a genetic hitch affecting the ALK protein. Researchers are still probing its use in other cancers involving the c-Met protein
- Posted from my iPad
Location:Georgetown Tx,United States
IPad HD