In the fight to defeat the most deadly forms of cancer, scientists in Omaha are enlisting a supercomputer instead of a robot.
To find a drug that could target troublesome cells, researchers formerly had to test compounds solely in a lab. Big pharmaceutical companies would keep hundreds of thousands of different compounds on hand, and a robotic device — doing work that used to be performed by lab technicians — would combine the different compounds in huge trays.
"You spent years and literally millions to tens of millions (of dollars) on the front end," and it took months, said Nick Palermo, a computational biochemist from the University of Nebraska-Lincoln. Using the Holland Computing Center supercomputer at UNO's Peter Kiewit Institute, where Palermo works, that same review of all the compounds can be performed in days for much less money, he said.
Palermo is among the researchers from UNL, the University of Nebraska at Omaha and the University of Nebraska Medical Center who are funded
by a $250,000 grant from the Nebraska Research Initiative. Their goal is to identify the best drug candidates that could lead to better treatments of cancers, particularly pancreatic cancer.
Pancreatic cancer, although relatively rare (2.8 percent of all new cancer cases), is an especially deadly form of cancer: Just 6.7 percent of patients survive for five years or longer after being diagnosed, the National Cancer Institute says.
In 2006, Surinder Batra, a professor and chairman of the department of biochemistry and molecular biology at UNMC, and his colleagues reported the discovery of a gene that they called pancreatic differentiation factor 2, or PD2. It is found in embryos but also in people with pancreatic cancer, said Joe Runge, director of business development for UNeMed, UNMC's technology licensing arm.
PD2 helps embryos build functioning organs and tissues, but it can become a problem if it gives cancer cells the same set of growth instructions.
"One property that's sort of common amongst tumors and embryonic cells is self-renewal," Runge said. "They're able to very aggressively divide and continue to produce.
"In pancreatic and ovarian cancer, the reason that tumors don't respond to treatment is that there is this subpopulation of cells that are just very, very resistant; very, very difficult to kill."
Batra found that PD2 is more highly expressed in those cells, Runge said.
UNeMed officials talked to pharmaceutical companies that are focused on finding new drugs to treat such cancer cells. A number of them were interested in the PD2 target, Runge said, "but for them just to invest at that stage is too early — they view it as too risky." So Batra and his colleagues need a list of compounds that shut down PD2, Runge said.
"If we can have something which can target cancer stem cells, it can prevent relapse, it can prevent drug resistance," Batra said.
That's where Palermo comes in. He studies the protein that the PD2 gene produces and has made a three-dimensional model of it. He then can simulate thousands of drug-molecule interactions.
Pointing to a program on his laptop, Palermo said, "Essentially, this turns molecules into a very complicated Lego set for me to play with."
"You're basically going to see a thing that looks like a blob," he said, dumbing it down for a reporter. "There are pockets in that blob. Those are the places where we're going to say, 'Hey, a drug might fit here and interact with it.' So what we get from the modeling is seeing where are we actually going to target a drug."
Of the money provided by the Nebraska Research Initiative, $50,000 went to computer hardware that has been added to the Kiewit Institute's supercomputer.
Palermo said the computer will "throw about 7 million or so compounds" at the PD2 protein. Through criteria that he has established, "we narrow it down to 200 or so," he said. "And I'll recommend, from those, maybe 20 or 30 to the folks at the med center and go, 'Let's buy these first, let's test these on the cells first. Because these make the cut.'"
Jonathan Vennerstrom, a professor in the UNMC College of Pharmacy, will review the recommendations and guide efforts away from results that may be impossible or too toxic for further testing. Batra and others then will test the compounds in a lab and on animal models.
"My point of being here is to give them the best possible starting point," Palermo said.
Officials say the collaboration could serve as a blueprint for future drug development research at the University of Nebraska.