Genetic and gene expression diversity among maize inbreds

Abstract

Alterations in the regulation of gene expression have long been proposed to be crucial in contributing to phenotypic variation. To systematically compare variation in the expression of transcription factors and other genes across genotypes and tissues, the transcriptomes of 5 tissues were deeply sequenced from 27 inbred maize lines, comprising the founders of the Nested Association Mapping (NAM) population. Consistent with the hypothesis that variation in gene regulation contributes to phenotypic diversity, the genes that exhibit the most variable expression (G-VE genes) across genotypes are enriched in transcription factors. A proportion of those G-VE genes also demonstrated extreme expression variation across tissues. We further identified approximately 5 million SNPs and InDels from the RNA-Seq dataset, and determined the minimum numbers of haplotypes per gene in the NAM founders. Consistent with the complementation model of heterosis, 1,319 Filtered Gene Set (FGS) genes are affected by Large Effect SNPs (LES), which altered start codons, stop codons or splice sites. 26% of these LES alleles are rare alleles (with allele frequencies < 0.1). However, an additional 44% of LES has allele frequencies more than 0.8, possibly due to inaccurate annotation of genes or unidentified isoforms in the B73 reference genome. Both CNVs and gene haplotypes were found to have tissue-specific contributions to variation in gene expression. Our analysis revealed both regulatory and genetic variants’ roles in contributing to phenotypic variation.