A systematic survey of human tissue-specific gene expression and splicing reveals new opportunities for therapeutic target identification and evaluation

Abstract

Differences in the expression of genes and their splice isoforms across human tissues are fundamental factors to consider for therapeutic target evaluation. To this end, we conducted a transcriptome-wide survey of tissue-specific gene expression and splicing events in the unprecedented collection of 8527 high-quality RNA-seq samples from the GTEx project, covering 36 human peripheral tissues and 13 brain subregions. We derived a weighted tissue-specificity scoring scheme accounting for the similarity of related tissues and inherent variability across individual samples. We showed that ~50.6% of all annotated human genes show tissue-specific expression, including many low abundance transcripts vastly underestimated by previous array-based expression atlases. As utilities for drug discovery, we demonstrated that tissue-specificity is a highly desirable attribute of validated drug targets and tissue-specificity can be used to prioritize disease-associated genes from genome-wide association studies (GWAS). Using brain striatum-specific gene expression as an example, we provided a template to leverage tissue-specific gene expression to identify novel therapeutic targets. Mining of tissue-specific splicing further reveals new opportunities for tissue-specific targeting. Thus, the high quality transcriptome atlas provided by the GTEx is an invaluable resource for drug discovery and systematic analysis anchored on the human tissue specific gene expression provides a promising avenue to identify novel therapeutic target hypotheses.

A limited set of transcriptional programs define major cell types

“…We used the patterns of expression of cell-type-specific genes to estimate the cellular composition of human tissues and organs from GTEx bulk tissue transcriptome data (version 6, 8555 samples, 31 tissues, 544 individuals) (The GTEx Consortium 2017). We used xCell (Aran et al 2017), using the sets of genes specific to epithelial, endothelial, and mesenchymal major cell types derived from EN-CODE, and specific to brain (neural) and blood derived from GTEx (Yang et al 2018) as signatures, and computed the enrichments of these cell types in each GTEx tissue sample (Supplemental Methods 8).…”

Methylation Quantitative Trait Loci are Largely Consistent across Disease States in Crohn’s disease

“…Our findings establish that blood mQTLs are robust and reproducible, supporting their generalizability across age groups, ancestries, disease status, and DNA sample source, and offer a valuable source for future studies examining blood mQTLs. Recent studies have identified different sets of differential DNAm patterns in various tissue types [21,22], and previous epigenomic and transcriptomic studies in IBD suggest that disease-relevant tissue samples are needed to develop precise biomarkers for the disease subtypes [5,[23][24][25]. Our findings show that among genome-wide-significant mQTLs, genetic influences on DNAm are generally consistent across blood and ileum, even among IBD SNPs and CD-associated CpG sites.…”

Modeling Pathway Dynamics of the Skeletal Muscle Response to Intravenous Methylprednisolone (MPL) Administration in Rats: Dosing and Tissue Effects

“…Tissue-specificity is a recognized yet underutilized resource particularly in drug discovery ( Yang et al, 2018 ; Yao et al, 2018 ; Ryaboshapkina and Hammar, 2019 ). The lack of a detailed understanding of the underlying gene regulation is clearly a major roadblock ( Sonawane et al, 2017 ).…”

“…Tissue-specificity is a recognized yet underutilized resource particularly in drug discovery (Yang et al, 2018;Yao et al, 2018; Supplementary Table D. Ryaboshapkina and Hammar, 2019).…”