Correlation and cluster analysis of immunomodulatory drugs based on cytokine profiles. (February 2018)

• Record Type: Journal Article
• Authors: Wallner, Fredrik K; Hultquist Hopkins, Malin; Woodworth, Nina; Lindvall Bark, Therese; Olofsson, Peter; Tilevik, Andreas
• Is Part Of: Pharmacological research. Volume 128(2018)
• Page Start: 244
• Page End: 251
• Publication Date: 2018-02
• Subjects: Immunosuppressive -- Clustering -- PCA -- Cytokines; Pharmacology -- Periodicals; Pharmacology -- Periodicals; Research -- Periodicals; Médicaments -- Recherche -- Périodiques; Pharmacologie -- Périodiques; 615.105
• DOI: 10.1016/j.phrs.2017.10.012
• Languages: English
• ISSNs: 1043-6618

Abstract

Graphical abstract: Abstract: Drug discovery is a constant struggle to overcome hurdles posed by the complexity of biological systems. One of these hurdles is to find and understand the molecular target and the biological mechanism of action. Although the molecular target has been determined, the true biological effect may be unforeseen also for well-established drugs. Hence, there is a need for novel ways to increase the knowledge of the biological effects of drugs in the developmental process. In this study, we have determined cytokine profiles for 26 non-biological immunomodulatory drugs or drug candidates and used these profiles to cluster the compounds according to their effect in a preclinical ex vivo culture model of arthritis. This allows for prediction of functions and drug target of a novel drug candidate based on profiles obtained in this study. Results from the study showed that the JAK inhibitors tofacitinib and ruxolitinib formed a robust cluster and were found to have a distinct cytokine profile compared to the other drugs. Another robust cluster included the calcineurin inhibitors cyclosporine A and tacrolimus and the protein kinase inhibitors fostamatinib disodium and sotrastaurin acetate, which caused a strong overall inhibition of the cytokine production. The results of this methodology indicate that cytokine profiles can be used to provide a fingerprint-like identification of a drug as a tool to benchmark novel drugs and to improve descriptions of mode of action.