PGRN Network-wide Project: Transcriptome Analysis of Pharmacogenes in Human Tissues. C. E. French1, A. Chhibber2, E. R. Gamazon3, S. W. Yee2, X. Qin4, E. Theusch5, A. Webb6, S. T. Weiss7,8, M. W. Medina5, E. G. Schuetz9, A. L. George10, R. M. Krauss5, C. Q. Simmons10, S. E. Scherer4, N. J. Cox3, K. M. Giacomini2, S. E. Brenner1 1) University of California, Berkeley, CA; 2) University of California, San Francisco, CA; 3) University of Chicago, Chicago, IL; 4) Baylor College of Medicine, Houston, TX; 5) Children's Hospital Oakland Research Institute, Oakland, CA; 6) Ohio State University, Columbus, OH; 7) Brigham and Women's Hospital, Boston, MA; 8) Harvard Medical School, Boston, MA; 9) St. Jude Children's Research Hospital, Memphis, TN; 10) Vanderbilt University, Nashville, TN.
Gene expression variation is crucial to the etiologies of common disorders and the molecular underpinnings of pharmacologic traits; however, the nature and extent of this variation remains poorly understood. The NIH Pharmacogenomics Research Network (PGRN) Network-wide RNA-seq project aims to create a community resource containing quantitative information on annotated and novel isoforms of genes involved in therapeutic and adverse drug response (pharmacogenes). Using 18 samples from each of 5 tissues of pharmacologic importance (liver, kidney, adipose, heart, and lymphoblastoid cell lines [LCLs]), we performed transcriptome profiling by RNA-Seq with the goal of determining differences in expression of pharmacogenes across tissues and between individuals. The data were analyzed for expression quantification, and we used the JuncBASE tool developed by members of our consortium to identify and quantify splicing events. In each of the tissues and LCLs, 11,223-15,416 genes were expressed at a substantial level. In pairwise comparisons of tissues, 105-211 pharmacogenes were differentially expressed (2-fold difference, FDR<0.1). For example, as expected, the CYP enzymes CYP2C19 (MIM 124020) and CYP2D6 (MIM 124030) were 10-fold and 100-fold more highly expressed in the liver than in other tissues. Other important drug metabolizing enzymes such as DPYD (MIM 612779) and TPMT (MIM 187680) showed more balanced gene expression patterns across the tissues. We observed that 72-93% of pharmacogenes are alternatively spliced within each tissue. There was substantial variation in both annotated and novel splicing events both between tissues and between individual samples of the same tissue. For example in SLC22A7 (MIM 604995), a gene encoding a transporter for various drugs, we found evidence of a novel alternative last exon that is variably spliced between individuals. LCLs are important pre-clinical models for human genetic studies, but they highly express less than half of pharmacogenes as compared with the 66-83% expressed at a substantial level in each of the physiological tissues. However, a number of genes like BRCA2 (MIM 600185) and SLC6A4 (MIM 182138) are much higher in LCLs than the tissues, as are alternative splice events of many genes. In conclusion, these studies demonstrate that important pharmacogenes are variably expressed across tissues of pharmacologic relevance, and across different individuals, and that the vast majority are alternatively spliced.
You may contact the first author (during and after the meeting) at