Skip to Content

XSEDE Science Successes

« Back

Probing DNA for Cancer Therapies


Published on July 28, 2016 by Makeda Easter

There are myriad examples throughout history where an error led to a discovery, a new invention, or a cure. Even on a biological level, errors are important — take DNA for example.

DNA (deoxyribonucleic acid) contains all of the genetic information for a living organization and is involved in many fundamental biological processes. Information in DNA is stored in chemical bases that pair up. The order of these base pairs contain essential information for an organism. Physical properties of DNA, such as elasticity, strength, and elongation, also play an important role in cell interactions.

The smallest error, such as a molecule inserted between two neighboring base pairs — called an intercalator — can change the structure of DNA and lead to genetic abnormalities that cause cell death. Because cancer is a disease caused by the uncontrollable division of cells, many anti-cancer drugs exploit this biological error.

"The main difference between the different types of chemotherapy drugs is related to the targets," said Arman Fathizadeh, physics researcher at the University of Illinois at Chicago (UIC). "For example, intercalating drugs target cells. But other drugs target enzymes which act on DNA."


According to the researchers' estimates, simulations are about five to ten times faster on Stampede than local resources at UIC.
Fathizadeh is among a group of researchers with Fatemeh Khalili-Araghi at UIC investigating intercalators and their effect on DNA in collaboration with Boise State University's Physics Department and Micron School of Materials Science and Engineering. The groups are looking into the effects of the drug doxorubicin used in cancer chemotherapy. Doxorubicin primarily treats leukemia and cancer of the bladder, breast, and stomach.


"We need to shed light on intercalating drugs and DNA," Fathizadeh said. "I believe this study will make a contribution to new cancer therapies by helping people to design novel drugs for several types of cancer.

To investigate these micro-level changes in DNA, the researchers are employing nanotechnology and advanced computing resources. Nanoscale technologies, specifically nanopores, allow the researchers to dive deep into the biological processes that cannot be seen with a regular microscope.


Arman Fathizadeh, postdoctoral research associate, Department of Physics, University of Illinois at Chicago
"While much emphasis has been placed on solid-state nanopore technologies as a next generation DNA sequencing technology, their more immediate impact may be in probing tertiary structure of single molecules. This has important implications for development of cancer therapeutics with reduced cardiotoxicity," said David Estrada, director of Boise State University's Integrated NanoMaterials Laboratory and coordinator for the Micron School of Materials Science and Engineering Graduate programs.




Each pore is comprised of a single layer of molybdenum disulfide with a solution of doxorubicin. As researchers apply voltage to both sides of the pore, ions move along an electric field. Inside each nanopore solution is DNA so that researchers can study the interactions of DNA with atomically thin materials. The translocation of DNA blocks the pore for ions and causes a drop in ionic current.

"Measuring the current drop can provide some information about the shape, configuration, and size of the DNA," Fathizadeh said. "And we can do this experiment and simulation with and without intercalators to see how the molecules affect DNA conformation and translocation dynamics.

Fathizadeh performs this process with molecular dynamics simulations which allows the team to capture details that they aren't able to see with experimentation. Through simulations, the researchers can better understand the effects of intercalators.

To perform these simulations, Fathizadeh originally used the local computing resources at UIC. But the researchers wanted to scale up their simulations and model the translocation process for much longer DNA that mirror actual systems in experiments.

"We have a much bigger system consisting of one and a half million atoms. We are also going to simulate the process for much longer time scales," Fathizadeh said.


Schematic representation of a supercoiled DNA ring translocating from a molybdenum disulfide nanopore.
Following advice from his advisor, Khalili-Araghi, Fathizadeh turned to the NSF-funded Extreme Science and Engineering Discovery Environment (XSEDE). Through XSEDE, the researchers received access to several advanced computing resources— Stampede at the Texas Advanced Computing Center (TACC), Comet at the San Diego Supercomputer Center, and Bridges at the Pittsburgh Supercomputing Center to perform molecular dynamics.


"XSEDE provides very fast computing resources and processors and we can use those compute nodes for parallel processing," Fathizadeh said. "At the same time, XSEDE resources are extremely efficient for the type of molecule dynamic simulations we use. We did the benchmarking on these machines and observed that they are working much faster on solutions compared to our local resources."

According to the researchers' estimates, simulations are about five to 10 times faster on Stampede than local resources at UIC.


"XSEDE resources are extremely efficient for type of molecule dynamic simulation we use. We did the benchmarking on these machines and observed that they are working much faster on solutions compared to our local resources."
Arman Fathizadeh, postdoctoral research associate, University of Illinois at Chicago
"So far we studied intercalating ethidium bromide and learned that this particular intercalator can only affect the translocation time. DNA translocation time through nanopore just slows down, but there are no effects on the ionic currents," Fathizadeh said. "Fortunately, our results were in line with the experiments at Boise State University. Now we're extending our research to doxorubicin and translocation of DNA supercoils. We hope that this research will help us to understand the fundamental mechanisms of this cancer drug on DNA structure and cell death."


"We performed our simulations 10 times faster using XSEDE resources, which helped us to simulate translocation of a supercoiled DNA from a nanopore," said Fathizadeh. "Hopefully, this can help us bridge our simulations with our experiments to understand the effect of doxorubicin on DNA." 

In the coming months, the researchers will continue to perform these simulations and are expecting to publish the first part of their research on ethidium bromide, an intercalating agent commonly used as a fluorescent tag in molecular biology laboratories.