A New Strategy for Identifying Mechanisms of Drug-drug Interaction Using Transcriptome Analysis: Compound Kushen Injection as a Proof of Principle

Hanyuan Shen,Jian Cui,Wei Wang,Yuka Harata-Lee,Zhipeng Qu,R Daniel Kortschak,David L Adelson,Thazin Nwe Aung

doi:10.1038/s41598-019-52375-3

Abstract

Drug-drug interactions (DDIs), especially with herbal medicines, are complex, making it difficult to identify potential molecular mechanisms and targets. We introduce a workflow to carry out DDI research using transcriptome analysis and interactions of a complex herbal mixture, Compound Kushen Injection (CKI), with cancer chemotherapy drugs, as a proof of principle. Using CKI combined with doxorubicin or 5-Fu on cancer cells as a model, we found that CKI enhanced the cytotoxic effects of doxorubicin on A431 cells while protecting MDA-MB-231 cells treated with 5-Fu. We generated and analysed transcriptome data from cells treated with single treatments or combined treatments and our analysis showed that opposite directions of regulation for pathways related to DNA synthesis and metabolism which appeared to be the main reason for different effects of CKI when used in combination with chemotherapy drugs. We also found that pathways related to organic biosynthetic and metabolic processes might be potential targets for CKI when interacting with doxorubicin and 5-Fu. Through co-expression analysis correlated with phenotype results, we selected the MYD88 gene as a candidate major regulator for validation as a proof of concept for our approach. Inhibition of MYD88 reduced antagonistic cytotoxic effects between CKI and 5-Fu, indicating that MYD88 is an important gene in the DDI mechanism between CKI and chemotherapy drugs. These findings demonstrate that our pipeline is effective for the application of transcriptome analysis to the study of DDIs in order to identify candidate mechanisms and potential targets.

Highlights

Www.nature.com/scientificreports medicines, especially traditional Chinese medicines (TCM) which are usually made from several herbs, can exert their effects on conventional medicines both through pharmacokinetic and pharmacodynamic effects
High-throughput omics-related techniques have been widely used for identifying novel disease biomarkers or potential drug targets[16,17], very limited research has applied them to the investigation of Drug-drug interactions (DDIs)
Because transcriptome based approaches generate very large data sets, we adopted a hierarchical approach for our analysis of DDIs between Compound Kushen Injection (CKI) and chemotherapy drugs

Summary

Introduction

Www.nature.com/scientificreports medicines, especially traditional Chinese medicines (TCM) which are usually made from several herbs, can exert their effects on conventional medicines both through pharmacokinetic and pharmacodynamic effects. In order to better understand DDIs and deal with the difficulties caused by complex components in herb-drug interactions, we propose a pipeline to apply transcriptome analysis for the study of DDIs. CKI was used as a test drug in combination with different chemotherapy agents, and was found to have different effects on cancer cells when combined with doxorubicin or 5-Fu (synergistic with doxorubicin and antagonistic with 5-Fu). We have identified hundreds of differentially expressed genes that are correlated with opposite effects of CKI and chemotherapy agents on cell viability or apoptosis These genes indicate that several cancer related pathways, such as DNA replication and cell cycle, are perturbed differently by CKI under different medical circumstances. Compared to previous DDI studies focused on transporters, metabolizing enzymes and therapy targets, our methods can provide a comprehensive and deeper analysis of interactions, that may help to pinpoint potential therapeutic or side effects, and explain the mechanisms underlying DDIs

Methods

Results

Conclusion