(Inside Science) — Researchers have looked back in time a billion years to solve a mystery that could lead to cures for some cancers a few years in the future.
In a paper printed in the February 20 issue of the journal Science, scientists at Brandeis University in Waltham, Massachusetts, described why a drug known to attack several forms of cancer works so well and why it fails with certain people. It could help drug makers eventually design pharmaceuticals for specific groups of patients, or even individuals.
Current chemotherapy for cancer involves essentially injecting poison into a patient's body, which not only kills the cancer cells, but raises havoc in normal ones nearby. The treatment is at best inefficient and often horrible for patients.
The Brandeis research centers on the drug Gleevec, about the closest thing to a "magic" bullet cancer drug modern medicine has found. Gleevec, also marketed as Glivec or Imatinib, has essentially cured most patients with a rare form of cancer called chronic myeloid leukemia and has been used for colon cancer and others originating in the gastric system.
It became somewhat controversial when a group of more than 100 doctors protested the high cost of the drug — about $100,000 a year. The company that sells Gleevec, Novartis, responded that the prices is so high because it works so well, curing 80 percent of patients.
By why doesn't it work on everybody?
Chronic myeloid leukemia is caused by a malfunction deep in the human genome. Something in chromosome 9 switches places with something in 22 and that shortens 22 and produces an unusually long chromosome 9, sometimes called the Philadelphia chromosome after the city in which it was discovered.
The Philadelphia chromosome produces a fusion gene called BCR-ABL which produces a protein that makes blood cells go wild. The mystery has been why Gleevec is so specific and only shuts down the bad proteins and not any of the closely related ones. This is the key to its success: it leaves healthy cells alone, minimizing side effects.
BCR is almost identical to the protein produced by ABL. Unexpectedly, Gleevec works especially well against production of ABL's protein but doesn't work at all with the closely-related BCR protein. They differed by 146 amino acids or proteins, about half.
Led by Dorothee Kern at Brandeis and the Howard Hughes Medical Foundation, the researchers decided to find when the two proteins evolved from a common ancestor, something never done before.
"It's not like looking for fossils," she said.
They found the answer in the family tree, recording a divergence that occurred about a billion years ago. The two proteins evolved through time, becoming more complex, Kern said. But they had a common ancestor, now called ANC-AS, a primitive protein found in a single-cell organism a billion years ago.
Throughout the course of evolutionary history, genes get duplicated, said Ryan Gutenkunst, a professor of molecular and cellular biology at the University of Arizona in Tuscon.
"There are several ways it can go," he said. "It can just duplicate itself, which is the more boring outcome. Or they can begin to diverge." One becomes different and the other remains the same; or they can both diverge. In the case reported by Kern, they diverged.
"The natural evolutionary pressure on the [proteins] is more complexity," Kern said. "It is why you are more complex than a yeast cell. When ANC-AS was the only version there were just a few kinases [a kind of protein]. Humans now have more than 500. The difference is in signaling processes, how they get turned on and off."
The trick for pharmaceutical companies is to produce compounds that bind with "the screwed up" proteins, but not the healthy 500 found in humans, she said. Gleevec does that because it only binds to ABL and that means fewer side effects, she said.
Currently, cancer patients can have their genomes profiled, a relatively new process brought about by dramatic decreases in price. These profiles are essentially searches for sick proteins within the genome.
"They sequence the entire genome to see which is the sick one," Kern said.
The goal is then to find a drug that will aim for just that protein. It would kill those drugs without slaughtering everything around them.
"If you have cancer you have 10 proteins that mutated for leukemia or lung cancer. Some patients have protein A; some patients have protein B, but there are many patients who all have cancer because of protein A," she said. In that hypothetical new drugs could be aimed at all those patients with protein A.