The Philadelphia Chromosome

by Jessica Wapner

  A year or two after Matter and Lydon began working together, Schering opened up a laboratory in Dardilly, a rural village in the south of France. The lab, with about sixty-five people, had been created specifically to study interferon and other so-called immunostimulants. Matter was asked to serve as the director, and Lydon went with him. But shifting finances resulting from changes in the political dynamics of France at the time, coupled with Schering-Plough’s simultaneous acquisition of the California-based company Dinex, which had been focusing on molecular immunology, dimmed the promise of the Dardilly lab. In a very short time it became obsolete, shrinking to twenty employees. At the same time, Matter was getting fed up with the bureaucracy of the sprawling company. He had supervisors in Paris, New Jersey, and San Diego, and he constantly had to make the rounds to keep them informed and appeased. Eventually, the situation became explosive. In the early 1980s, Matter left the company on unfriendly terms.

  He accepted a job at Ciba-Geigy, a pharmaceutical giant situated along the Rhine River in Basel, Switzerland. Despite the administrative difficulties he’d had at Schering, the years there had left him inspired about creating pharmaceuticals for cancer. He’d been following the studies of oncogenes and kinases avidly, and he wanted to turn those discoveries into drugs that would improve the outlook for cancer patients. It was the same dream that would bring Druker to Nashoba Community Hospital just a couple of years later.

  When Matter was still at Schering, he and Lydon spoke frequently about kinases, how they seemed like the perfect drug target. By the time Matter arrived at Ciba-Geigy, links between oncogenes and kinases were cropping up in several cancer research labs. It was far more than src at this point. A cancer-causing protein known as v-erbB turned out to be related to a kinase called epidermal growth factor receptor, or EGFR. Another, v-sis, was connected to a kinase called platelet-derived growth factor receptor, or PDGFR. In the late 1970s, a Japanese researcher named Yasutomi Nishizuka discovered an enzyme family called protein kinase C, or PKC. In 1982, Nishizuka and a researcher from France named Monique Castagna discovered that phorbol esters, compounds derived from a natural substance in plants and known to cause skin cancer in mice, target PKC to exert their malignant tendencies.

  There was more. As investigators began peering inside tumor cells, again and again they found excess amounts of EGFR, PDGFR, and PKC. The phenomenon came to be known as “overexpression”; genes expressed proteins, and so an excess amount of protein was considered an overexpression. The presence of abnormally high quantities of these kinases strengthened the theory that they were not just linked to cancer but were somehow responsible for it. It was like finding the leaky faucet responsible for a persistent dripping noise. EGFR was found in lung and brain cancers. Excess amounts of PDGFR had been found in several solid tumors and some rare blood cancers.

  Then, of course, Matter and Lydon were also seeing the reports on the connection between the Philadelphia chromosome and the Bcr/Abl tyrosine kinase, the fusion protein that resulted when the genes bcr and abl were brought next to each other in the translocation. Nearly all people with CML had this genetic mutation that created the out-of-control kinase that led to leukemia. Among the growing body of evidence linking kinases to cancer, this one was the best established at the time. People weren’t saying that the so-called deregulated kinase caused CML yet, but many were thinking it. Like Druker, Matter and Lydon understood that kinases were being revealed as a driving mechanism of cancer.

  At Ciba-Geigy, Matter wanted to start a kinase inhibitor program, a research project dedicated solely to creating compounds “to be used as pharmacological means to get rid of cancer via dysregulated kinases,” recalls Matter. Like Lydon, who had stayed at Schering after he left, Matter was convinced that kinases were the perfect drug target.

  Matter would need help, though, and who better than Lydon? He had accrued years of knowledge about kinases ever since he’d befriended the postdocs in Phil Cohen’s lab, and he was inspired about the possibility of exploiting them to thwart cancer.

  In the summer of 1984, right around the time he was reading Tony Hunter’s Scientific American article “The Proteins of Oncogenes,” Lydon received a call from Matter offering him a job with his new kinase inhibitor program at Ciba-Geigy. Lydon wasted no time in making his way to Basel. “He was very courageous,” says Matter. “I didn’t have a lab, I didn’t have technicians. I had nothing.”

  Matter presented a proposal for a kinase drug program to his supervisors at Ciba-Geigy. The company agreed to fund the project, though begrudgingly. Just a year earlier, Ciba-Geigy had sworn off cancer therapeutics and was refusing to pour any money down what it considered to be an expensive and fruitless drain. But Matter’s boss, who was also a good friend, had assured Matter that he could build a cancer portfolio when he came to work there. The best way to do that, Matter was advised, was to keep the program very, very small. The boss also made sure Matter included multiple therapeutic categories in his research department. Alongside the anti-kinase drugs, Matter worked on aromatase inhibitors and phosphonates, two important classes of drugs today. The kinase program was hidden in plain sight, attracting little attention for the moment.

  The tectonic plates of cancer care were creaking into motion. The proposal approved by Ciba-Geigy outlined one of the first efforts to design a drug against a specific, well-known target, a sharp contrast to the shot-in-the-dark, trial-and-error approach that characterized the bulk of cancer treatment history. Defiant against the incremental improvements that each new chemotherapy drug had brought, Matter’s vision was to make a truly meaningful leap in cancer care by focusing on the kinase. “Today we know [kinases] are so important in the regulation of the growth of . . . cells. It’s common knowledge; you learn it in the introduction to cancer biology,” said Jürg Zimmermann, a chemist who would soon join Matter’s team. “But at that time, there were only a few pioneers who really thought that drug discovery should look into the effect kinases have on the growth of cancer cells.” Matter was one of those pioneers.

  The approach was called “rational drug design,” and the promise of it was huge. As scientists dived into the molecular biology of cancer and began surfacing with concrete results, the notion of fashioning drugs against specific cellular targets materialized as a natural next step. Targeting the root cause of cancer could make for a far more effective drug than chemotherapy. Most chemotherapy drugs carpet-bombed the body in the hope of hitting the cancer cells. If they were aimed toward anything, it was the fastest growing cells in the body, which is why they often led to hair loss, the least debilitating but perhaps most iconic side effect. If individual targets could be identified for each type of cancer, then care could be shaped personally around each patient, side effects would diminish, and the ultimate outcome of treatment—a longer life without disease—would vastly improve.

  The moniker “rational drug design” stemmed from the sequence of events by which the medication was created. The end point—say, a haywire kinase—was already known, and the compound was being designed against that predefined result. The approach was the complete reverse of chemotherapy, in which the rationale was far more hypothetical, and it was possible only as a result of accruing evidence that cancer was caused by genetic abnormalities that led to changes in the cellular products of the affected genes. As oncogenes emerged, so did the idea that drugs could be designed specifically against their cancer-causing products in the cell.

  The theory behind kinase inhibition was straightforward. With rational drug design, it was the fit that mattered, creating a medication that was perfectly shaped for the surface of its target. Like all proteins, kinases are three-dimensional structures with bumpy surfaces. The ridges and dips on the surface provide cleavage sites, areas for the compound to grab onto, like a mountain climber finding a jutting rock for his next step. The bumpier the surface, the easier it would be to design a drug that would adhere to its target.

  No technology was available to see a
n actual kinase, but its structure could be deduced based on its chemical makeup. All the team had to do was make a compound that would fit onto the kinase, preventing it from binding ATP. Like a gloved hand fitting perfectly over a mouth to block the next breath, the drug would stop the runaway kinase in its tracks, thereby halting cancer progression. Kill the kinase and you kill the cancer.

  The approach had its skeptics. The idea of creating a molecule that blocked only one kinase at a time seemed ludicrous to many researchers around the world. If too many kinases were targeted at once, a person could easily die. This kind of selectivity in a drug, the hallmark of targeted therapy, had never even been attempted, so there was no foundation on which to build what seemed like a very faulty tower. And as logical as rational drug design sounded, it was still theoretical. No such drugs had yet been created.

  Matter and Lydon were undaunted. The rationale was solid. Plus the field was still relatively simple. In the mid-1980s, when they were getting the project under way, about twelve protein kinases had been found. Creating a compound to selectively block just one of those was sure to be difficult, but it didn’t seem impossible. By 1987, about 65 genes were known to encode kinases. Today, 500 kinases have been identified. If Lydon and Matter had known there would turn out to be so many kinases—making the idea of selective inhibition seem preposterous—they might never have started.

  Because skepticism loomed large inside the company, Matter had been told to risk-balance his portfolio, to make sure that he was not betting the farm on something everyone else thought would fail. “Marketing had told him that with cisplatin, you could make millions, so what the heck [was he] doing?” Zimmermann said. But Matter, who’d already earned a reputation at Ciba for being headstrong and combative, would not be swayed. Cisplatin was a cytotoxic agent, a drug that poisoned cells throughout the body. It and drugs like it had made cancer treatments notorious for their wretched side effects. Matter and his team were trying to break out of the chemotherapy stronghold, not strengthen it. “His stubbornness allowed him to move forward and follow his mission,” said Zimmermann. “He wanted to change the way we did drug discovery in oncology.”

  By late 1984, evidence was mounting that Matter and his team really were onto something. A Japanese researcher named Hiroyoshi Hidaka had found that a group of compounds called isoquinolinesulfonamides could inhibit kinases. One of his molecules was particularly active against PKC, a kinase that prior research had shown to be overexpressed in some cancers, for example. Ultimately, Hidaka’s compounds would not turn out to be effective drugs against cancer (though one was eventually approved in Japan for the treatment of other conditions), but those first reports stirred interest. There were still more skeptics than cheerleaders, but the idea was getting some traction.

  BY LATE 1985, Matter’s group was growing. Peter Traxler, a chemist who’d been researching antibiotics at Ciba-Geigy since 1973, was moved to the project when the company decided to stop its antibiotics program. The biologists worked under Lydon’s supervision, and the chemists under Traxler’s. The most noteworthy other additions were Elisabeth Buchdunger, who worked with Lydon, and Zimmermann, who worked with Traxler. A chemist named Thomas Mayer joined the group and worked with Zimmermann on creating the first “hits,” compounds that showed some anti-kinase activity, and a scientist named Helmut Mett also came on board. Matter gradually inherited other Ciba employees who, like Traxler, were being moved off of other projects. They were a patchwork of researchers who together created a team that was low-key enough to not attract too much attention from the higher-ups at Ciba.

  They were also getting help from experts outside of Ciba. By the early 1980s, Basel had become a hive of biomedical research, in part because of the collection of talent housed at the Friedrich Miescher Institute, or FMI. The facility had been created in the 1970s by then-separate pharmaceutical companies Ciba and Geigy, but it operated independently of industry, doing basic research that could then inform the lab work and administrative decisions at the corporations. Not long after arriving in Basel, Lydon found out that Brian Hemmings, a former postdoc from Phil Cohen’s lab at Dundee with whom Lydon had mulled over kinases at the local pub, was now a scientist at FMI. Years later, with the scenery slightly changed, they resumed the conversation. “We spent a lot of time over beers discussing [kinases],” said Lydon, “including would Abl be a good target.”

  16

  _______

  A MACHINE WITH A VIRUS FOR A MOTOR

  Because this rational design approach to drug development was entirely new, there were no methods in place for doing it. So Lydon’s first task was to create a method for examining the compounds created by Ciba’s chemists. It was well and good to create a molecular structure that one thought would fit onto the kinase inside the cell. But with no way to test whether it would do anything to cells in a laboratory setting, what was the point? The team needed a way to screen compounds for activity against the stated drug target. That target could be any kinase that was implicated in the development of cancer or other serious diseases. This screening method had to work with whatever experimental molecule they threw at it, and it needed to work with all of the kinases that held promise as potential drug targets.

  The techniques available at the time, like gel electrophoresis, the tool Owen Witte had used to separate the Gag/Abl protein into its constituent molecules, were arduous and unpleasant. Today, hundreds of experimental compounds can be checked very quickly for activity against numerous targets at once. Any time a new potential target is found inside tumor cells, drug companies can use high-throughput screening to scan their libraries of experimental compounds for hits. A plate with hundreds to thousands of tiny wells in which a drug target and experimental compound are mixed can be tested by a computer that measures properties that are altered when the target is reacting to a drug. For example, a program might measure how reflective a protein is, because reflectivity increases when a protein is bound by a drug. But back when Lydon and Matter were starting out, there was no technology or protocol for testing drug candidates against a specific target because drugs had never been designed this way before. They had to invent a way to do it.

  The challenge was producing large amounts of whatever protein they were trying to target. They needed the protein in large amounts, and it had to be active, still functioning as it would inside the body. Only by having the actual kinase available in cell culture could they determine whether one of the chemists’ creations stopped it from phosphorylating protein. Those tests had to be done on isolated enzymes before they could be done on actual cells. But active kinase, the thing that supposedly launched the cascade of events that led to cancer, existed only inside cells. The problem was similar to the one Naomi Rosenberg had solved with her transformation system, bringing the inner, molecular world of the cell into the exposed world of the laboratory. Just as Rosenberg had found a way to study mouse cells outside of a mouse, Lydon and Matter needed to find a way to produce a high-quality enzyme outside of a cell that could be used day in, day out for as long as it took to test anti-enzyme activity of their experimental molecules.

  And that was when the trouble began. “It was painstakingly slow to develop this active enzyme,” Matter said. “Everything went wrong that could go wrong.”

  Eventually, the group succeeded in producing live, active kinase using Escherichia coli, the bacteria better known as a feared food contaminant. The assay was such a breakthrough that Lydon published the method, a rare move for someone in industry, where secrecy is the default setting.

  But the breakthrough was a step, not a leap. The problem with the E. coli assay was that it only worked with the Abl kinase. The well-defined link between Abl and CML made this kinase an attractive drug target; there was little question about its importance in the development of the cancer. But that didn’t necessarily make Abl attractive to a pharmaceutical company. The problem was that CML was a rare disease. With an annual incidence of about 5,000 people per year in t
he United States and between approximately 70,000 and 140,000 people worldwide (one to two per 100,000 people), CML was relatively insignificant to a large pharmaceutical company. A drug to treat this rare cancer would have an extremely small “market,” as the patient population is referred to. The company needed to be sure that whatever medication eventually came out of this program would reach as many people as possible. A drug for a more common cancer would also be far more lucrative.

  For those reasons, kinases that had been found in more common cancers held much more interest than Abl. Three other kinases that had also made their appearance in the scientific literature—PKC, PDGFR, and EGFR—were turning up everywhere. PDGFR was expressed in almost every type of cancer. EGFR had also been unearthed in major cancer types, including lung cancer, a disease with an annual incidence of more than a million people. PKC and EGFR had been found in breast cancer. By the mid-1980s, breast cancer diagnoses had climbed to 350 per 100,000 women in the United States, with about 1.5 million diagnoses worldwide. What’s more, emerging cardiology research was implicating PDGFR in a common complication following the insertion of balloons to treat blocked arteries. The idea of creating a kinase inhibitor for cancers that struck hundreds of thousands of people per year, and maybe even for heart disease, captivated the imagination of the research team and the business planners.