NEW YORK (GenomeWeb News) – A pair of new studies is demonstrating the potential applications of genomics to help find new therapeutic applications for existing drugs.
Stanford University systems biology researcher Atul Butte and his colleagues developed a computational strategy for comparing shifts in gene expression that are characteristic of different disease states with the gene expression changes that occur when cell lines are treated with various drugs. By looking for situations in which these expression signatures were anti-correlated, showing opposing patterns in response to a given drug and disease state, the team was able to make predictions about which compounds might effectively treat certain conditions.
"If a drug exerts a change on gene-activity pattern that is opposite to that exerted by a disease, then that drug may have a therapeutic effect on that disease," Butte said in a statement.
In two papers, both appearing online today in Science Translational Medicine, research teams led by Butte illustrated how this genomics-based drug repositioning strategy could be used to find new uses for existing drugs: the anticonvulsant topiramate and the anti-ulcer drug cimetidine.
For one of the papers, researchers brought together publicly available expression data for inflammatory bowel disease samples and gene expression data linked to 164 drug treatments. In the process, they found computational clues that topiramate, which is typically used to treat epilepsy, seizures, and migraines, might be useful for treating IBD. An existing IBD treatment, prednisolone, was also quite high on the list of candidate compounds they identified.
Indeed, the team's follow-up studies in a rat model of the IBD colitis (induced in the animals using trinitrobenzenesulfonic acid) indicated that the topiramate could curb disease symptoms and pathology.
"These findings suggest that topiramate might serve as a therapeutic option for IBD in humans," Butte and his co-authors wrote, "and support the use of public molecular data and computational approaches to discover new therapeutic options for disease."
In the other paper, researchers used a similar but more large-scale approach to look for drug-disease pairs using gene expression data for 164 compounds as well as microarray data for 100 diseases included in the National Center for Biotechnology Information's Gene Expression Omnibus.
"Instead of examining a single drug-disease pair or even looking at reactions of a large set of drugs on a single disease, we focused on discovering connections between drugs and diseases across all the available gene measurements," the study authors explained. "[W]e present evidence that this strategy can confirm already known therapeutic uses for drugs and uncover new uses for [US Food and Drug Administration]-approved drugs."
Among the 53 significant drug-disease interactions they detected, the team found evidence that the ulcer drug cimetidine might be effective against lung adenocarcinoma — a prediction that was supported by their subsequent experiments, including experiments in which they used cimetidine to treat xenograft mouse models of the lung cancer.
Moreover, they reported, several of the other drug-disease interactions that they found seem to be consistent with known therapeutic patterns. For instance, the approach paired several anticancer compounds known as HDCA inhibitors with three types of brain tumors that are known to show some response to such drugs.
Despite their encouraging results so far, study authors cautioned that the new computational method relies on having accurate gene expression profiles for candidate compounds and diseases, which are not always available. And, while their experiments in animal models appear promising, they noted that more pre-clinical testing and clinical trials of cimetidine are needed to determine whether it has the same success for treating lung cancer in humans.
"This work is still at an early stage," National Institutes of Health Pharmacogenomics Research Network Director Rochelle Long, who was not involved in the studies, said in a statement, "but it is a promising proof-of-principle for a creative, fast and affordable approach to discovering new uses for drugs we already have in our therapeutic arsenal."
"Bringing a new drug to market typically takes about $1 billion, and many years of research and development," Long said. "If we can find ways to repurpose drugs that are already approved, we could improve treatments and save both time and money."
In an accompanying perspectives article in Science Translational Medicine, genetic medicine, genomics and systems biology researcher Yves Lussier, and hematology and oncology researcher James Chen, both from the University of Chicago, said the studies mark a move away from reductionist biology and towards a more multifaceted, or so-called "scalar," view.
"The basis of this shift toward computational integrative approaches," they wrote, "has strong precedence in scalar theories of biological information — or the idea that, as we transition among biological levels, unique properties appear and disappear — and could prove useful in future efforts for drug repurposing."
"The two articles from the Butte group move us toward more encompassing and more integrative approaches for drug repurposing," they wrote. In addition, they noted, similar expression anti-correlation strategies might also prove useful for other research problems as well.
"Clearly, numerous other intermediary genome-wide metrics could be measured at other scales of biology, from microRNA signatures to comparative genomic hybridization data," Lussier and Chen concluded. "A conceptual framework for guiding the analysis of these scales' inputs and outputs is, therefore, of paramount importance for their accurate interpretation and application."