The Philadelphia Chromosome

by Jessica Wapner

  The challenge lay in more than creating the perfect shape. To be a good inhibitor, the compound also had to stick to the kinase. Creating that snug fit between a molecule and the enzyme was an achievement, but it was no guarantee that the molecule would stay there.

  For his molecule to be a viable candidate for a drug, that bond had to be strong. “The stronger the bond, the less of the substance you have to administer to the patient later on,” he explained. A compound that was weak might work only at an impossibly high dose. A compound that adhered well to the kinase—that was potent, in other words—would work at a lower dose.

  plate 15

  When the chemists managed to create a compound that was both selective—adhering to one specific kinase only—and potent, they sent it to Elisabeth Buchdunger and the other biologists who were part of the team. The biologists had to test each candidate to see if it was active in cancer cells. A molecule might fit into the ATP binding site of one kinase and not any others, and it might stick like cement, but ultimately that meant nothing. The compound had to result in the death of cancer cells; otherwise it was worthless. To become a drug candidate, a compound had to be selective, potent, and active. For the first few years of the program, molecules were handed back and forth, from the chemists to the biologists and back again, each round a new attempt at achieving all three qualities—each time, a moment of starting over, adjusting, and trying again.

  Not surprisingly, Buchdunger’s tests usually rejected the candidate. A successful compound had to break through the membrane of a cell, make its way through the cytoplasm inside, find the target kinase, attach to the kinase and stay there long enough to kill the cell. Month after month, Zimmermann or the other chemists sent potent and selective compounds to Buchdunger only to discover that they weren’t active. Adjusting the molecule so that it might kill the cell often led to a loss in potency or selectivity, and they would have to start all over again.

  Among the already existing chemicals the chemists looked to was one called 2-phenylaminopyrimidine, a compound known for its anti-inflammatory effects. When Zimmermann and Mayer tested it against PKC, one of the three kinases in which the company was most interested, the enzyme was blocked. But the effect was too weak; a drug version of the 2-phenylaminopyrimidine would have required too large a dose—grams, rather than the usual milligrams—to be practicable. The body cannot handle such high quantities of a powerful chemical; and even if that were not the case, the administration of a ridiculously large pill or lengthy infusion would be impossible, even parceled out over a day. But having seen that the chemical did something, the team deemed this the “lead” compound, the one it should focus on.

  Lead in hand, the chemists had to find a way to improve it. Adding a molecule called 3’-pyridyl to the original chemical scaffolding heightened its activity. With the introduction of a single six-sided molecule, suddenly Buchdunger could see that the compound was blocking PKC with much greater efficiency. Next the chemists added a benzamide group, a molecule created by exposing a version of benzoyl (a component of the acne-fighting combination benzoyl peroxide) to ammonia, and sent the molecule back to the biologists. This time, the molecule was even more active against multiple tyrosine kinases, including PDGFR and Abl.

  Then the experiment took a surprising turn. The chemists introduced a methyl group, a combination of carbon and hydrogen, to the growing molecule. In fact, it was just a portion of a methyl group, a so-called flag methyl, which the chemists stuck onto an open slot in the middle of the original backbone. In the diagram of its chemical structure, the molecule now had a single wayward line jutting out from one corner of one hexagon. That single scrap of chemical transformed the entire molecule.

  When Buchdunger screened the compound again, she noticed that it no longer inhibited PKC. The flag methyl had eliminated that effect. Before the flag methyl was attached, there were a few different ways in which the atom could form into a single cohesive molecule—several conformations, in chemistry terms. The flag methyl group forced the compound into one immovable arrangement, like a host finalizing the seating for a dinner party. And in that inflexible setup, the molecule couldn’t bind to PKC anymore.

  The compound was still active—incredibly so. But now that strong activity was directed against Abl. The candidate had been their lead because of its activity against PKC. Now it had become the potent tyrosine kinase inhibitor the team had dreamed of creating—against Abl, the CML-driving kinase. “Just by doing a minor change in the structure, the activity of the compound changed from a PKC to an Abl inhibitor,” said Traxler. The compound also blocked PDGFR, though with less strength.

  The project’s requirements were satisfied: ­­The experimental molecule was selective for a particular kinase, it was a potent inhibitor of that specific kinase, and it was active against cells. Yet the team wasn’t quite sure what to make of this new chemical. Here they had designed a compound that exactly matched the one they’d envisioned. But it was for the wrong kinase. Now what? “We’ve got a kinase, we’ve got an inhibitor,” Brian Hemmings recalled saying when discussing the results with the team. “All we need now is a disease.”

  The timing was extraordinary. It was 1990, and the final proof that the Philadelphia chromosome, and it alone, caused CML had been made. Druker, Lydon, Matter, and others had already accepted this notion after the evidence had amassed several years earlier. But in strict scientific terms, the mutation had not been proved to be the single trigger for CML.

  A man named George Daley, yet another member of the Baltimore lab, had finally accomplished this remaining feat. Daley took one group of mice and filled their marrow with the mutant bcr/abl gene present in the Philadelphia chromosome. Next, he destroyed the bone marrow of a second group of mice with radiation. He injected the second group of mice with the marrow from the first group, and the second group of mice developed CML. The experiment established the mutant chromosome, and therefore its protein product, Bcr/Abl, as the sole cause of CML.

  The proof bolstered Lydon’s belief that the kinase program at Ciba-Geigy should make Abl its top priority. Despite his and Druker’s conviction that Bcr/Abl was the best target for proving the principle of kinase inhibition, the program had maintained a general focus on all cancer-associated kinases. The enzymes that were considered to be more desirable targets were those associated with diseases far more common than CML. Now the scientific literature and the company’s efforts had converged, with Bcr/Abl as their meeting point.

  “That was exactly the time when Nick [Lydon] came and said, ‘Look, Bcr/Abl kinase is a very hot item,’” Traxler recalls. “So we changed from a PKC project to a Bcr/Abl project.” Lydon knew that CML was a rare disease and that its rareness made an anti-Abl drug less desirable to a pharmaceutical company. But he also knew that inhibiting Abl was the company’s best shot at proving the principle of kinase inhibition as a treatment for cancer. Because Bcr/Abl was solely responsible for CML, this cancer provided an ideal way to test out the idea. If a patient’s CML stopped progressing as a result of taking an Abl inhibitor, then the company could be assured that the drug was responsible for that change because there were no other cancer-causing factors at play. CML was the perfect testing ground for kinase inhibition, for rational drug design, for treating cancer as a genetic disease.

  Now Lydon had the perfect disease for proving the principle of kinase inhibition, and the perfect molecule. The chemists made one last addition to the anti-Abl molecule they’d created. A molecule called N-methylpiperazine improved the compound’s water solubility, turning it into a medicine that could be taken by mouth. They sent it to Buchdunger to test for activity.

  It worked. At last, about six years after Ciba-Geigy had green-lighted Matter’s proposal for a kinase drug development program, Buchdunger was able to report that the compound was potent, selective, and cellularly active. The final molecular formula was C29H31N7O • CH4SO3. Described another way, its designation was 4-[(4-methyl-1-piperazinyl)methyl]-N-[4
-methyl-3-[[4-(3-pyridinyl)-2-pyrimidinyl]amino]-phenyl]benzamide methanesulfonate. It was a white to off-white to brownish powder with a molecular mass of 589.7; it also carried the weight of forty-three years of science history. The compound was named CGP-57148B. “To me,” said Buchdunger, “it was already quite a little bit of a miracle.”

  19

  _______

  TWO ENDINGS

  In the late 1980s, as the chemistry team in the Ciba-Geigy oncology research department was piecing together molecules like so many Legos, Brian Druker was still plodding away with his polyomavirus project. The work had gone from an interesting way to learn about molecular biology and oncogenes to an increasingly dull project with no end in sight. He’d had a handful of publications come out of his work, but nothing all that gripping. “I was five years in the lab,” said Druker. “I didn’t have a lot to show for it.” By 1990, just when he was getting married and settling down in a house in the Boston suburbs, he was starting to wonder what he was doing with his life.

  He’d kept up with his weekly clinic at Nashoba Community Hospital, and he knew that ultimately he wanted to help patients. That had been the goal of his immersion in basic science all along, however unarticulated it had been. By 1990, as he took stock of his talents and interests, he saw his growing expertise in kinase biology and his expertise in cancer care finally coming together. “Why don’t I work on a human disease caused by kinase?” he asked himself.

  George Daley had just nailed down the final proof about the cause of CML. Because Druker and Lydon had struck up a friendship founded on their common interest in kinase inhibitors, and because he had an inkling that he might want to do research with such a compound, Druker had been following the Ciba-Geigy team’s progress. He knew that Lydon had included Abl as one of the kinases against which new molecules were screened. Now, it was Druker’s turn to make a bold move. After all the research, all the patients, all the talks with Nick Lydon, he could no longer stand to be on the sidelines. He wanted to be part of bringing this new kind of cancer treatment into the world. He wanted to make a tyrosine kinase inhibitor to treat CML.

  He paid a visit to Jim Griffin, a researcher whose lab was one floor below Tom Roberts’s lab, where Druker was working. Griffin was a myeloid biologist and knew the ins and outs of bone marrow and the cancers that grew there. The two of them had already worked together uncovering kinases involved in bone marrow biology. Druker asked Griffin if he would collaborate with him on studying the connection between CML and the signaling pathway triggered by the Bcr/Abl kinase. Griffin agreed.

  Until 1990, Druker’s seizing of opportunities was of a more passive variety. If a door was open, he’d walk through, but he wouldn’t force one ajar. For him, the decision to turn his attention exclusively to CML was different. “That was really the first time in my whole life that I had made a proactive decision,” he said. He collected cell samples from CML patients. He looked for phosphorylated proteins in those cells. He charted the signaling pathways activated in CML. “That’s what I want to do,” he realized. “I’ll figure out later where it all leads.” Having never worked with an experimental drug before, he had just a vague idea of how to start. He knew the Ciba-Geigy team was on its way to generating some compounds that might block the Bcr/Abl kinase. Perhaps he could test them on cells from actual CML patients.

  Just a few months after making this declaration to himself, his plans were squashed. Dana-Farber made an agreement with Sandoz, another pharmaceutical behemoth whose headquarters were on the opposite side of the Rhine from Ciba-Geigy. The companies were rivals in the development of drugs for all manner of illnesses. The agreement gave Sandoz exclusive access to the work going on at several Dana-Farber labs, including the one belonging to Tom Roberts, where Druker was still working.

  The arrangement was an increasingly common one between academia and industry. Within the universities, grant-funded research was unraveling the intricacies of cancer, heart disease, reproduction, neurological disorders, psychiatric disorders, sleep, allergies, and on and on. But there would always be a limit on how far an academic researcher could take the work because universities didn’t have the resources to, say, turn a new finding into a drug development program. Pharmaceutical companies were designed for just that purpose, but often lacked the raw material of original research that could be refined into a new medication. So contractual agreements sprouted up, giving industry access to the latest promising discoveries and giving universities another way to earn money. But the arrangement only worked if the relevant labs were forbidden from conferring with a company’s competitors. And though such agreements were on the rise, the offer from Sandoz to Dana-Farber was enormous at the time: $10 million per year for ten years. Overnight, Druker was cut off from communications with Lydon, Matter, and Buchdunger. “We could no longer pursue that relationship,” said Druker.

  Druker and Griffin decided to write a grant application to Sandoz, outlining how they would test a drug that inhibited the Bcr/Abl kinase. They showed, step by step, how they would take an anti-kinase drug that had proved worthy in the lab and investigate its effects on people. But Sandoz wasn’t interested in kinase inhibitors. They would give Druker and Griffin some money to develop tools to study them just in case the company ever took an interest. The company was lukewarm at best about the potential for these drugs to work and turn a profit. Druker and Griffin took whatever money the company was willing to give them to explore how, exactly, they would study a kinase inhibitor, were they to ever have one available to test. For Druker, the early 1990s were dedicated to further parsing of the signaling pathway that flowed downstream from the mutant Bcr/Abl tyrosine kinase. But because of Dana-Farber’s contract with Sandoz, accessing the inhibitors being created at Ciba-Geigy was out of the question.

  It was Druker’s personal life that came to inform his next steps. In 1992, he and his wife divorced. After two years of marriage, he’d been forced to face the truth that his work mattered more to him than his relationship. “I was not the best husband,” said Druker. “I got married because I thought it was the thing to do, and [she] was a lovely young woman. But when push came to shove, it was about my work, and so I wasn’t very present.”

  As he and his ex-wife headed in separate directions, he gradually rediscovered his love of the outdoors, something he missed from his San Diego days. He remained in Newton, a few miles outside of Boston, and, following the lead of his athletic roommate, another doctor, he began biking to Dana-Farber each day. The exercise quickly became a form of stress relief, a way to rid himself of the mental buildup of weeks of painstaking and often unsuccessful experiments. Soon biking became integral to surviving the lab, the frustrations, and the continual grief over his patients.

  His weekly clinic work continued, spanning a total of seven years. He treated people with all different types of cancer, with all the regimens still focused around chemotherapy, turning frequently to the towering experts at Dana-Farber, essentially the birthplace of modern cancer care, for information on prostate cancer, lung cancer, lymphoma. For the moment, the treatments had stagnated. The breast cancer drug Herceptin, directed against a genetic mutation present in some women, was not yet in clinical trials. Tamoxifen, which blocked estrogen and is today considered the first targeted therapy, was in widespread use. Estrogen, a hormone associated with female sexual traits, had long been associated with a certain variety of breast cancer. Thwarting the excess production of the hormone in patients with that type of breast cancer proved to be an effective treatment. But tamoxifen, though aimed against a specific chemical in the body, didn’t tackle cancer at its roots. Estrogen is a hormone, not a genetic mutation. A drug inhibiting the production of estrogen could stop or diminish breast cancer growth by removing this chemical on which it depended, but the underlying cause—whatever it was that triggered the excess production in the first place—was left untouched. It was like stopping a car by removing the gas pedal instead of the motor. (The drug, though highly be
neficial for many patients, also has some severe side effects, including a heightened risk of uterine and endometrial cancer.) In the eyes of doctors, patients, drug makers, and the public at large, cancer had not yet become a genetic disease. Medicine had not yet become personal to the degree of examining an individual’s DNA and tailoring treatment accordingly. The principle had not yet been proved.

  20

  _______

  GETTING OUT OF BOSTON

  By 1993, six years after arriving at Dana-Farber, Druker was beginning to feel restless and worried. The link between Bcr/Abl and CML was now firmly established and widely known. Around the world, new kinases were continually being discovered. Investigations into the molecular biology of cancer churned out an increasing number of connections to mutant versions of these enzymes. New oncogenes routinely appeared in the literature. One day, Druker would have an idea about how to target a kinase with a drug, and the next day, he’d read about that same idea in the scientific literature. He began to feel trapped by his circumstances. “If I have a good idea I should be able to execute it and be the one that publishes it,” he thought. “But I can’t do that if I’m the only person doing the work.” Griffin had proved to be an excellent collaborator, but he didn’t have his mind set on creating a CML drug as fiercely as Druker did.

  It was time to start his own lab where he could steer the research and focus exclusively on the development of a kinase-blocking drug for CML. And there was only one way to do that: He had to ask Dana-Farber for an assistant professor position, some funding, and some space. “And that didn’t go very well,” said Druker.