Skip to main content
Skip to main content

Comprehensive analysis of transcriptional response upon multiple drug treatments

Life Sciences


Investigating the drug response at the transcriptional level is powerful for understanding the mechanism of drug adverse effects, elucidating inter-individual drug response variability and developing personalized therapies. However, drug perturbation experiments are limited to immortal cell lines which do not reflect the true physiology and gene expression of the liver, a key organ for drug metabolism. This liver is hard to access in humans and multiple drug treatments to determine gene expression would be deemed unethical for transcriptomic profiling. Moreover, available resources are mostly based on microarray which lacks the broad dynamic range on RNA-sequencing data in capturing transcript abundance. We treated 68 primary human hepatocytes derived from donors of African American ancestry with 6 known inducers of drug metabolism, Omeprazole, Phenobarbital, Dexamethasone, Carbamazepine, Phenytoin and Rifampicine and measured the transcriptome via RNA-seq technology before and after treatment. This unprecedented resource allow us to characterize the drug-specific mechanism of action and identify generic signatures of drug response. We performed pair-wise differentially expressed (DE) gene analysis and identified thousands of genes that were up- or down-regulated upon drug treatment. Phenobarbital and Omeprazole demonstrated strongest response and Carbamazepine demonstrated lowest response in terms of the number of DE genes. Common overexpressed DE genes across drug treatments are enriched for targets of a transcriptional factor, TCF11, which suggests its role as an essential gene in modulating drug response. Grade of membership (GoM) clustering captures the heterogeneity among drug profiles. Phenobarbital-specific cluster highlights the ATP synthesis process. Both DE gene analysis and GoM clustering demonstrated the similarity between Phenytoin and Carbamazepine, which may result from their common therapeutic application on epilepsy. In conclusion, analysis across drug transcriptomic profiles holds the potential to provide insights into the commonalities and differences of mechanisms of genetic regulation of drug response.

Yizhen Zhong, et al.


April, 2018

DOI: 10.21985/N2VX0K