Regulators of expression in Zea mays

Abstract

To assess the impact of regulatory variation on a genomic scale, a large scale gene expression resource was created from seven tissues profiled using 3’ RNAseq in a diverse set of nearly 300 maize lines. This resource enabled testing associations between multiple levels of biological organization from genotype to intermediate phenotype to terminal phenotypes including yield. Specifically, associations were tested between genotypes and gene expression (eQTL) and between gene expression and downstream phenotypes. This large-scale expression resource was used to quantify the impact of rare allele abundance on expression and to assess the relationship between dysregulation of expression and terminal fitness phenotypes. Expression extremes were found to be strongly associated with local rare allele abundance. Additionally, formerly rare alleles whose frequencies increased as a result of the tropical-temperate bottleneck were found to have an outsize impact on expression in predominantly temperate germplasm. Furthermore, deviation of expression was found to be predictive of seed weight, highlighting the importance of gene regulation for fitness and yield. A second study assessed the distribution of eQTL relative to their regulated genes and found considerable evidence of both tissue-specific and shared cis regulation. The study also revealed that long distance cis regulation at the scale of tens of kilobases is not an exception, with more than 20% of significant primary cis eQTL mapping more than 20kb from the regulated gene. The common variant eQTL were also found to be enriched among known functional annotations, suggesting the eQTL are also functional. A third study examined the capacity to associate variability in expression directly with terminal phenotypes of interest as a supplement linking genotype to phenotype. This transcriptome-wide association study (TWAS) is a useful accessory to genotype based associations. In a combined test with genome-wide association results, TWAS improves the capacity to re-detect genes known to underlie oligogenic kernel phenotypes. This improves not only the capacity to link genes to phenotypes, but also underlines the widespread importance of regulation for phenotype.