IU School of Medicine researchers dive deep into DNA to predict adverse drug reactions

  • Oct. 25, 2016

INDIANAPOLIS – Two Indiana University School of Medicine scientists are exploring uncharted genetic pathways in search of tools to predict whether patients will react well or poorly to drugs, in research supported by a new $1.69 million, three-year grant from the National Cancer Institute.

Yunlong Liu, PhD, associate professor of medical and molecular genetics and of biostatistics, and Todd Skaar, PhD, associate professor of medicine, are focusing on differences in people's reactions to chemotherapy drugs paclitaxel and clofarabine, but say they expect their techniques will be applicable to assessing other drugs' toxicities as well.

Drs. Liu and Skaar were among five research teams to receive grants in a program overseen by the National Human Genome Research Institute, all of which were designed to evaluate genetic variants that have been discovered in less-traveled sections of the genome. Most researchers have focused on the sections of the DNA that are directly involved in the process of producing proteins, Liu said. But those sections of DNA – the gene coding regions – make up only about 2 percent of human DNA, he said.

Other scientists receiving the recent NHGRI grants will focus on portions of the DNA that work as regulatory mechanisms, such as controlling when the coding regions turn on and off.

"Our interests are more in the mechanics of splicing," said Dr. Skaar.

Splicing refers to a process – in the gene-coding regions – of removing sections of DNA that aren't needed for protein expression. Those unneeded sections are called introns. The sections of DNA needed are called exons. After the cell's molecular machinery connects together the appropriate exons, the protein construction process continues.

But how to tell exons from introns? With little DNA "signs." Between the exons and introns are short sections of DNA code that identify adjacent sections as one or the other, much like road signs that tell drivers what's ahead. Much as road signs that are missing or askew can result in lost drivers, variations in the DNA signs can result in proteins that malfunction, or don't function.

Skaar and Liu are collaborating with Michael Eadon, assistant professor of medicine at IU School of Medicine, and with Eileen Dolan, PhD, professor of medicine at the University of Chicago, who developed a list of genetic variants potentially related to drug toxicity by performing an analysis of a group of immune system cell lines in the laboratory.

Individually testing each of the 100,000 candidates in cells would be an impossible task, so Liu will use biostatistical techniques and computer machine learning to predict which of the candidates are most likely to affect drug toxicity.

Skaar's lab will then test the resulting 1,000 or so candidate variations to determine their activities in cells, using techniques his lab has developed to dramatically speed up the process.

The final candidate markers will be tested in Dr. Dolan's cell lines in Chicago, a process that should lead to clinical tests in humans to validate whether or not they can be used to accurately predict patients' responses to the chemotherapy drugs.

"Later, we expect that this process will be available to be used in literally every other genomic study that is out there," said Dr. Liu.

Todd Skaar, PhD, (L) and Yunlong Liu, PhD

Todd Skaar, PhD, (L) and Yunlong Liu, PhD

Print Quality Photo

Eric Schoch