Science Editor

The success of bortezomib has shown without a doubt that inhibiting the proteasome is a way to selectively kill cancer cells. Most recently, Millennium Pharmaceuticals Inc. presented top-line results in September showing that the drug leads a "highly statistically significant improvement" in all efficacy measures, including time to disease progression, complete remission rate, progression-free survival and overall survival, when combined with melphalan and prednisone as first-line treatment.

Detailed results from the study will be presented at next week's American Society of Hematology meeting in Atlanta. In the meantime, the company is preparing to file a supplemental new drug application for the compound by the first quarter of 2008. (See BioWorld Today, Sept. 18, 2007.)

But why the drug is selective is not obvious, given that its mechanism is anything but. It controls the lifespan of nearly 80 percent of proteins in normal cells. The proteasome "degrades a lot of proteins, so when it is inhibited, a lot of proteins are going to accumulate in the cell," Maria Soengas told BioWorld Today. But only cancer cells are likely to react by throwing in the towel.

In a paper in the Dec. 4, 2007, issue of the Proceedings of the National Academy of Sciences, senior author Soengas and her colleagues at the University of Michigan described experiments that suggested the reason for that specificity lies in the interplay between two particular proteins: the proapoptotic protein NOXA and the oncogenic transcription factor c-myc.

When Soengas describes cancer cells, they sound sort of like the Batmobile, albeit after it's been hijacked by a villain. A cancer cell "has many accelerators," in the form of oncogenes which make it divide rapidly, she said. "And in addition to going fast, it is sort of bulletproof" - in other words, because the accelerators are redundant, putting a brake onto one of them doesn't necessarily do much to stop the cell's runaway division.

"That's why these supertargeted drugs don't work very well," Soengas said.

NOXA levels rise in cancer cells when the proteasome is inhibited. In their PNAS paper, Soengas and her team first demonstrated that in melanoma cells, new RNA synthesis is necessary for the rise in protein levels to occur, leading them to focus on transcription factors.

Several likely candidates, including p53, turned out to be unnecessary for that rise in NOXA levels. But when Soengas and her group focused their attention on c-myc, they found that the oncogene apparently can sow the seeds of its own destruction by increasing NOXA levels.

The scientists identified c-myc-binding sites that promote transcription of NOXA, and demonstrated that the depleting endogenous c-myc via RNA interference prevented the accumulation of NOXA in melanoma and some other types of cancer cells, and that increasing c-myc levels in normal cells made them sensitive to bortezomib, as well. Taken together, the results suggested that bortezomib leads to selective apoptosis in cancer cells by increasing NOXA levels via increased levels of c-myc.

Soengas was cautious when asked about the potential clinical significance of her findings: "We did not cure melanoma," she said. "I don't want to overinterpret these studies - we are not there yet." But, she said, the results suggested that upregulating the production of c-myc might be a way to "make bortezomib kill better."

And such upregulation can be achieved in a way that is specific to cancer cells, meaning that bortezomib's increased activity would not have to come at the expense of more severe toxicity.

"You can find a window in time and in dose where it is possible to induce myc in tumor cells and not in normal cells," she said. "That's the beauty of it."