Cancer research gets into the groove

By Leah Eisenstadt, Broad Communications, November 12th, 2009
NOTCHTargeting proteins that control the expression of DNA has been a difficult task.
Image courtesy of Eric Smith, Dana-Farber Cancer Institute

Uncovering the molecular basis of cancer can be a double-edged sword. With decades of research and recent advances in cancer genomics, physicians may have a better idea of what genetic mutations cause a patient’s cancer. But, in many cases, they have no way of targeting the root cause, resorting instead to chemotherapy that kills dividing cells indiscriminately and brings a host of dangerous side effects.

One reason for this frustrating situation is that the dysfunctional proteins driving many cancers are notoriously difficult to target with drugs. The out-of-control cell growth that defines cancer results from runaway growth genes activated by regulator proteins, known as transcription factors, that sit on DNA and turn genes on and off. Transcription factors are often mutated in cancer, but scientists have been largely unable to design or find drugs capable of blocking the proteins.

A creative method of targeting these gene regulators has recently been applied to cancer by a multi-institutional collaboration of researchers from the Broad Institute’s Chemical Biology Program, Dana-Farber Cancer Institute, Harvard University, and Brigham and Women’s Hospital. By combining advances in chemistry, chemical biology and genomics, the scientists have devised a way to generate unique molecules that target transcription factors, and can serve as both valuable research tools and prototypes for new cancer medicines. The new work, led in part by co-senior author Jay Bradner, also a chemical biologist and oncologist at the Dana-Farber Cancer Institute and an associate member of the Broad Institute of MIT and Harvard, appears in the November 12 issue of Nature.

A physician specializing in blood cancers and stem cell transplantation, Bradner knows first-hand the frustration felt in the clinic when treatment options are inadequate. In his practice, he sees patients with advanced blood cancers like T-cell acute lymphoblastic leukemia (T-ALL). “Unfortunately, by the time I see these patients, they’ve already failed to respond to conventional therapy,” said Bradner. “It’s in these patients where new therapies are most urgently needed.”

Bradner’s colleague Jon Aster, an associate professor in the Department of Pathology at Harvard Medical School and co-author of the new study, had discovered in 2004 that 50% of patients with T-ALL have mutations in a gene called NOTCH, originally named for its effect on the notched wings of fruit flies when altered. NOTCH helps cells throughout the body make decisions about what to become, e.g., T-cell or B-cell. If NOTCH is mutated and becomes too active, it can drive cancer.

Because the NOTCH protein is a transcription factor, there were no clear ways to target it molecularly. “Proteins that regulate gene activity usually come together like two noodles, with expansive interfaces, so getting a molecule to bind just right and block the interface is a real challenge,” said Bradner. Targeting proteins with drugs or molecules often requires some kind of groove or pocket in the target. A small molecule can nestle into this groove and block the protein from doing its job. Transcription factors have such large, smooth interfaces that they are often termed “undruggable,” at least using conventional techniques of screening for drugs.

In 2005, a crucial piece of evidence was made public at a scientific meeting in Atlanta. There, Bradner heard a lecture by Aster, who had recently collaborated with Steve Blacklow, a co-author of the new study, to uncover the crystal structure that NOTCH forms when it partners with DNA and two other proteins. For the first time, scientists and physicians could view the shape of this gene regulatory team and visualize how the molecules worked in concert to flip a genetic switch and drive cancer.

“The structure was beautiful,” said Bradner. Interestingly, it revealed a long, spring-like section of one protein, known as an alpha-helix, that sits in a groove in the NOTCH protein. Bradner wondered if something could be designed to fit in the groove and prevent this protein from teaming up with NOTCH, essentially blocking its function. In his meeting program booklet, he began sketching out inhibitory molecules shaped like the alpha-helix, speculating how important such a molecule would be to researchers working to reveal NOTCH’s secrets in the lab. He also wondered if those inhibitory molecules might reveal a good spot for drugs to bind.

The alpha-helix is a common shape formed by proteins and chemists can build it in the lab. But a short section of alpha-helix tends to unravel, reducing its ability to bind a target. Bradner needed a way to constrain the alpha-helix into a spring-like shape. Back in Boston, he contacted Greg Verdine, Erving Professor of Chemistry at Harvard University, director of the Chemical Biology Initiative at Dana-Farber Cancer Institute, and co-senior author of the new study, who had previously developed a method of “stapling” tiny bits of protein known as peptides in place using non-natural amino acids synthesized in the lab.

“Jay [Bradner] came back from the conference as said, ‘This is a slam dunk for your technology’,” Verdine recalled. Not only is the stapling method tailored to alpha-helices, but it achieves feats impossible with other molecules. “Just adding this very simple chemical modification causes a bunch of things to happen all at once,” he said. The method doesn’t simply enforce the helical shape between ends of the “staple,” it holds the entire peptide tightly in shape. In turn, this prevents cellular enzymes from breaking down the proteins, a major route of peptide inactivation. Stapled proteins also stay active longer in the body, giving them a greater chance of performing their task. Most dramatically, stapling allows the peptides to enter cells. “That’s magical, because it solves the biggest problem we face with unmodified peptides,” said Verdine.

Ray Moellering, a graduate student in Harvard University’s Department of Chemistry and Chemical Biology who works with both Verdine and Bradner, created six overlapping stapled alpha-helices that covered the entire length of one of NOTCH’s partner proteins. One peptide, which they called SAHM1, was particularly promising.

They synthesized a fluorescent version of SAHM1 and proved that it did indeed penetrate cells — frequently a challenge for peptides. The peptide killed human cancer cells driven by NOTCH, and left alone those not driven by NOTCH — a promising result. Using gene expression microarrays and the Broad’s Gene Set Expression Analysis (GSEA) software, the scientists found that the peptide completely shuts down the NOTCH target genes, the very genes that cause cancerous cell growth.

Realizing that SAHM1 could accomplish such a feat — inhibition of a transcription factor in cells — the researchers wondered if the peptide they had designed to reveal where to put a drug, could perhaps be a drug itself. To test this hypothesis, they teamed up with Gary Gilliland, Professor of Medicine at Brigham and Women’s Hospital, and his graduate student, Melanie Cornejo, to develop a model of T-ALL in mice. Through clever work with luminescent firefly genes and bone marrow transplantation, they raised a fleet of mice with “glowing” tumors driven by the NOTCH mutation. In all the mice treated with SAHM1, NOTCH was inhibited and tumors shrank.

With a wealth of evidence in hand from years of work, Bradner and his team are now working to identify a collaborator in industry that may be able to develop this molecule into a drug. A directed inhibitor of NOTCH that works in humans could be useful not only for treating T-ALL, but other cancers like ovarian, breast, and lung cancer, in addition to other medical conditions such as atherosclerosis and immune diseases. In addition, Verdine and his fellow researchers are now building stapled peptides that inhibit the other three versions of NOTCH. They are also working to target other transcription factors, many thought to be undruggable, with stapled peptides. With the new method, he said, “it’s now open season for targeting transcription factors.”

The work, while not ready for use in patients, has implications beyond developing a new research tool and prototype drug for targeting NOTCH. It provides a generic method for targeting other transcription factors, a previously unattainable goal. Moellering explained, "Our approach could offer a template for targeting many other master regulators in cancer." Bradner added, “Our hope is that scientists will use this template strategy to create powerful tools for studying gene regulation throughout cell biology.”

This work was supported by the Leukemia and Lymphoma Society, the American Association for Cancer Research, the American Society of Hematology, the Harvard & Dana-Farber Program in Cancer Chemical Biology, the NIH and other funding organizations.

Paper(s) cited: 
Moellering et al. Direct inhibition of the NOTCH transcription factor complex. Nature DOI: 10.1038/nature08543