Network integration and modelling of dynamic drug responses at multi-omics levels

Nathalie Selevsek,Steven Niederer,Chris Bauer,Jos Kleinjans,Ralph Schlapbach,Stefano Gotta,Matthias Lienhard,Carla Pluess,Sally J Deeb ,Witold Wolski,Irina Agarkova,Johannes Schuchhardt,Jort J Merken ,Stefan Boerno,Gal Barel,Laura Kunz,Ines Smit,Alex Lewalle,Jasmine Minguet,Nicolas Bosc,Christoph Thiel,Fiona Hunter ,Vanessa Baier,Anne Hersey,Olivia Clayton,Bernd Timmermann,Florian Caiment,Ramona Nudischer,Timo Wittenberger,Bernardo Lino De Oliveira ,Lars Kuepfer,Marcha Verheijen,Henrik Cordes,Adrian Roth,Stéphane Heymans ,Hans Gmuender,Ralf Herwig,Yannick Schrooders,Ugis Sarkans,Alexandra Zerck,Francis Atkinson ,Patrick Guye,Ivo Bachmann

doi:10.1038/s42003-020-01302-8

Abstract

Uncovering cellular responses from heterogeneous genomic data is crucial for molecular medicine in particular for drug safety. This can be realized by integrating the molecular activities in networks of interacting proteins. As proof-of-concept we challenge network modeling with time-resolved proteome, transcriptome and methylome measurements in iPSC-derived human 3D cardiac microtissues to elucidate adverse mechanisms of anthracycline cardiotoxicity measured with four different drugs (doxorubicin, epirubicin, idarubicin and daunorubicin). Dynamic molecular analysis at in vivo drug exposure levels reveal a network of 175 disease-associated proteins and identify common modules of anthracycline cardiotoxicity in vitro, related to mitochondrial and sarcomere function as well as remodeling of extracellular matrix. These in vitro-identified modules are transferable and are evaluated with biopsies of cardiomyopathy patients. This to our knowledge most comprehensive study on anthracycline cardiotoxicity demonstrates a reproducible workflow for molecular medicine and serves as a template for detecting adverse drug responses from complex omics data.

Highlights

Uncovering cellular responses from heterogeneous genomic data is crucial for molecular medicine in particular for drug safety
A true integrative understanding of molecular mechanisms of adverse drug reactions can only be accomplished by interrogating different levels of molecular activities in a dynamic fashion, and this approach requires (i) capturing of the dynamic responses across time and dose in the system under study enabling dynamic network inference and quantitative modeling[14,15], (ii) identification of mechanistic networks of interacting proteins that are responsible for the systemic drug response[16,17], and (iii) validation of these networks in human patients
We demonstrate that proteins of the AC-induced cardiotoxicity (ACT) network were (i) clinically transferable to patients suffering from drug-induced cardiomyopathies and (ii) physiologically relevant as tested with a previously defined model for predicting ACT of the cardiomyocyte mitochondrion[22] and that network-based data integration has major potential for advancing the field of molecular medicine

Summary

Introduction

Uncovering cellular responses from heterogeneous genomic data is crucial for molecular medicine in particular for drug safety. Crucial for realizing this goal is a comprehensive characterization of adverse pathways across multiple regulatory systems In this context the identification of mechanisms through application of multiple-omics technologies that enable capturing the full width of molecular responses upon drug treatment combined with integrative data analysis approaches has been defined as the key strategy for solving the task[1]. We demonstrate that proteins of the ACT network were (i) clinically transferable to patients suffering from drug-induced cardiomyopathies and (ii) physiologically relevant as tested with a previously defined model for predicting ACT of the cardiomyocyte mitochondrion[22] and that network-based data integration has major potential for advancing the field of molecular medicine

Methods

Results

Conclusion