A team of Columbia University Medical Center (CUMC) researchers have developed a computer algorithm that can provide scientists with a digital “window” through which to see how drugs produce pharmacological effects inside the body. The researchers maintain that results of their study could help scientists create more efficient drugs less prone to producing side-effects, suggest ways to regulate a drug’s activity, and to identify novel therapeutic uses for boyh new and existing pharmaceutical compounds.
“For the first time we can perform a genome-wide search to identify the entire set of proteins that play a role in a drug’s activity, says study co-author Dr. Andrea Califano, the Clyde and Helen Wu Professor of Chemical Systems Biology and chair of the Department of Systems Biology at CUMC.
The study, entitled “Elucidating Compound Mechanism of Action by Network Perturbation Analysis“, is published in the journal Cell (Cell, 2015; 162 (2): 441 DOI: 10.1016/j.cell.2015.05.056), and coauthored by Dr. Califano with Jung Hoon Woo, Yishai Shimoni, Wan Seok Yang, Prem Subramaniam, Archana Iyer, Paola Nicoletti, Mara Rodriguez Martinez, Gonzalo Lopez, Ronald Realubit, Charles Karan, Brent R. Stockwell, and Mukesh Bansal, (all at CUMC), and Michela Mattioli of the Fondazione Istituto Italiano di Tecnologia.
Images Caption: DeMAND – a method to predict genes involved in mechanism of action of a compound. DeMAND predictions can be used to identify compound similarity. Known MoA genes are identified with high precision, sensitivity, and specificity,Novel predictions of both MoA and similarity were experimentally validated (Image Credit: Califano lab/Columbia University Medical Center)
The coauthors note that genome-wide identification of small-molecule compounds’ mechanism of action (MoA), thus characterizing their targets, effectors, and activity modulators, represents a highly relevant yet elusive goal that has critical implications for assessment of a compound’s efficacy and toxicity. THey observe that current approaches are labor-intensive and largely limited to elucidating high-affinity binding target proteins.
In their study, the researchers introduce a regulatory network-based approach that assesses genome-wide MoA proteins based on the global dysregulation of their molecular interactions following perturbation by a focus compound. Analysis of cellular perturbation profiles identified established MoA proteins for 70 percent of compounds tested, and elucidated novel proteins that were experimentally validated.
Finally, unknown-MoA compound analysis revealed altretamine, an anticancer drug, to be an inhibitor of glutathione peroxidase 4 lipid repair activity — a process that was experimentally confirmed, thereby revealing unexpected similarity to the activity of the agent sulfasalazine. The researchers maintain this observation suggests that regulatory network analysis can provide valuable mechanistic insight into elucidation of small-molecule MoA and compound similarity.
A CUMC release explains that while scientists design drugs to pinpoint molecular targets in the cell, once enters the human body, a drug becomes part of an incredibly complex biochemical system, typically interacting with other molecules in ways difficult to predict. This undesirable chemical cross-talk causes side-effects that are frequently serous enough to prevent promising drug candidates from being approved for clinical care, and unfortunately, current experimental methods are not sophisticated enough to enable scientists to identify the full repertoire of proteins affected by a particular drug.
However, by utilizing a new approach for analyzing drug-induced changes in disease-specific patterns of gene expression, a new algorithm that members of Dr. Califano’s lab at CUMC have devised called DeMAND (Detecting Mechanism of Action by Network Dysregulation) identifies the genes involved in a particular drug’s effects — both wanted and unwanted. This method could help more precisely predict a drug’s effects, including undesirable off-target interactions, suggest ways of regulating a drug’s activity, and perhaps even identify novel therapeutic uses for FDA-approved drugs, three factors that pose critical challenges in drug development.
The DeMAND method involves creation of a computational model of the network of protein interactions that occur in a diseased cell. Experiments are then performed to track gene expression changes in diseased cells as they are exposed to a drug being investigated. The DeMAND algorithm combines data from the propositional model with data from the experiments, and is thereby able to identify the complement of proteins most affected by the drug.
The CUMC scientists observe that their DeMAND algorithm improves on more labor-intensive and less efficient investigational and experimental methods that are capable only of identifying targets to which a drug compound binds most strongly. Because the DeMAND algorithm is able to identify many molecules that are affected in addition to a drug’s direct target, it can provide a more comprehensive side-effect profile of the subject drug.
The researchers observe that thus far, DeMAND’s predictions are proving accurate when tested with follow-up experiments. They researchers report that for example when they exposed human diffuse B-cell lymphoma cells to a panel of drugs, the algorithm was able to identify 70 percent of previously documented targets. “The accuracy of the method has been the most surprising result,” says Dr. Califano.
The DeMAND algorithm therefore makes it possible to identify a variety of compounds that cause similar pharmacological outcomes. A practical example is that by using DeMAND, the researchers were able to demonstrate that a similar subset of proteins is affected by the unrelated drugs sulfasalazine and altretamine. Altretamine is currently approved for ovarian cancer, but the DeMAND results suggest that, like sulfasalazine, it could be effective if for bowel inflammation or rheumatoid arthritis as well.
The Cell paper’s co-senior author Mukesh Bansal sees great potential in this approach, observing that “DeMAND could accelerate the drug discovery process and reduce the cost of drug development by unraveling how new compounds work in the body. Our findings on altretamine also show that it can determine novel therapeutic applications for existing FDA-approved drugs.”
This research was supported by grants from the National Institutes of Health (5U01CA168426, 1U01CA164184-02, 3U01HL111566-02, 5U54CA121852-08, 5R01CA097061, R01CA161061), New York Stem Cell Science (C026715) and the Howard Hughes Medical Institute.
For more information, visit http://cumc.columbia.edu or http://columbiadoctors.org.
Sources:
Columbia University Medical Center
Cell
Image Credits:
Columbia University Medical Center
