Hidden Genetic Regulation of Human Complex Traits via Brain Isoforms

Lu Pan; Chenqing Zheng; Zhijian Yang; Yudi Pawitan; Trung Nghia Vu; Xia Shen

doi:10.1007/s43657-023-00100-6

Hidden Genetic Regulation of Human Complex Traits via Brain Isoforms

Phenomics. 2023 Mar 20;3(3):217-227. doi: 10.1007/s43657-023-00100-6. eCollection 2023 Jun.

Authors

Lu Pan^#^{1

2}, Chenqing Zheng^#¹, Zhijian Yang^#¹, Yudi Pawitan², Trung Nghia Vu², Xia Shen^{1

2

3

4

5}

Affiliations

¹ Biostatistics Group, School of Life Sciences, Sun Yat-Sen University, Guangzhou, 510006 China.
² Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, 17177 Sweden.
³ State Key Laboratory of Genetic Engineering, School of Life Sciences, Fudan University, Shanghai, 200433 China.
⁴ Center for Intelligent Medicine Research, Greater Bay Area Institute of Precision Medicine (Guangzhou), Fudan University, Guangzhou, 511458 China.
⁵ Centre for Global Health Research, Usher Institute, University of Edinburgh, Edinburgh, EH8 9AG UK.

^# Contributed equally.

Abstract

Alternative splicing exists in most multi-exonic genes, and exploring these complex alternative splicing events and their resultant isoform expressions is essential. However, it has become conventional that RNA sequencing results have often been summarized into gene-level expression counts mainly due to the multiple ambiguous mapping of reads at highly similar regions. Transcript-level quantification and interpretation are often overlooked, and biological interpretations are often deduced based on combined transcript information at the gene level. Here, for the most variable tissue of alternative splicing, the brain, we estimate isoform expressions in 1,191 samples collected by the Genotype-Tissue Expression (GTEx) Consortium using a powerful method that we previously developed. We perform genome-wide association scans on the isoform ratios per gene and identify isoform-ratio quantitative trait loci (irQTL), which could not be detected by studying gene-level expressions alone. By analyzing the genetic architecture of the irQTL, we show that isoform ratios regulate educational attainment via multiple tissues including the frontal cortex (BA9), cortex, cervical spinal cord, and hippocampus. These tissues are also associated with different neuro-related traits, including Alzheimer's or dementia, mood swings, sleep duration, alcohol intake, intelligence, anxiety or depression, etc. Mendelian randomization (MR) analysis revealed 1,139 pairs of isoforms and neuro-related traits with plausible causal relationships, showing much stronger causal effects than on general diseases measured in the UK Biobank (UKB). Our results highlight essential transcript-level biomarkers in the human brain for neuro-related complex traits and diseases, which could be missed by merely investigating overall gene expressions.

Supplementary information: The online version contains supplementary material available at 10.1007/s43657-023-00100-6.

Keywords: Alternative splicing; Expression quantitative trait loci (eQTL); Genome-wide Association Studies; Isoform-ratio quantitative trait loci (irQTL); Neuro-related human complex traits.