Single-Parent Expression Is a General Mechanism Driving Extensive Complementation of Non-syntenic Genes in Maize Hybrids

Abstract

Maize (Zea mays L.) displays an exceptional degree of structural genomic diversity [1, 2]. In addition, variation in gene expression further contributes to the extraordinary phenotypic diversity and plasticity of maize. This study provides a systematic investigation on how distantly related homozygous maize inbred lines affect the transcriptomic plasticity of their highly heterozygous F hybrids. The classical dominance model of heterosis explains the superiority of hybrid plants by the complementation of deleterious parental alleles by superior alleles of the second parent at many loci [3]. Genes active in one inbred line but inactive in another represent an extreme instance of allelic diversity defined as single-parent expression [4]. We observed on average ∼1,000 such genes in all inbred line combinations during primary root development. These genes consistently displayed expression complementation (i.e., activity) in their hybrid progeny. Consequently, extreme expression complementation is a general mechanism that results on average in ∼600 additionally active genes and their encoded biological functions in hybrids. The modern maize genome is complemented by a set of non-syntenic genes, which emerged after the separation of the maize and sorghum lineages and lack syntenic orthologs in any other grass species [5]. We demonstrated that non-syntenic genes are the driving force of gene expression complementation in hybrids. Among those, the highly diversified families of bZIP and bHLH transcription factors [6] are systematically overrepresented. In summary, extreme gene expression complementation extensively shapes the transcriptomic plasticity of maize hybrids and might therefore be one factor controlling the developmental plasticity of hybrids.

Highly Genotype- and Tissue-specific Single-Parent Expression Drives Dynamic Gene Expression Complementation in Maize Hybrids

“…These results indicate a vital role of constitutively expressed syntenic genes for the maintenance of basic cellular function. In contrast, we found PAV SPE and non-PAV SPE genes are highly enriched in the non-syntenic gene space and depleted for genes with possible functions from the curated gene set, consistent with previous findings (Swanson-Wagner et al, 2010;Paschold et al, 2014;Marcon et al, 2017;Baldauf et al, 2018).…”

“…In a previous study using maize root tissue of different developmental stages it was shown that hybrids consistently expressed more genes than inbred parents in all the investigated triplets (Paschold et al, 2012;Paschold et al, 2014;Baldauf et al, 2018). In this study, this same trend was observed for the vast majority of the root triplets (90%, 43 out of 48) ( Figure 3).…”

“…Transcriptome divergence between maize inbred parents and their F 1 -hybrid progeny has been previously investigated (Guo et al, 2004;Swanson-Wagner et al, 2006;Springer and Stupar, 2007b;Stupar et al, 2008;Paschold et al, 2012;Paschold et al, 2014;Baldauf et al, 2016;Marcon et al, 2017;Waters et al, 2017;Baldauf et al, 2018). While these studies provided insights into transcriptional variation between inbreds and hybrids and its potential role in heterosis, they were performed in one or a handful of parent-hybrid triplets or only a single tissue.…”

“…More recently, the role of single parent expression (SPE) genes in heterosis has been investigated (Paschold et al, 2012;Paschold et al, 2014;Baldauf et al, 2016;Marcon et al, 2017;Baldauf et al, 2018). SPE refers to a situation in which expression can only be detected in one of the two parental lines of a hybrid.…”

Non-Homology-Based Prediction of Gene Functions

“…The majority of these genes are evolutionarily young, with the evidence that the majority of them are conserved between maize and its wild progenitor teosinte, but only a small minority were found to be conserved at syntenic locations the genome of Sorghum bicolor a related species which diverged from the lineage leading to maize approximately 12 million years ago (Swigoňová et al, 2004;Sun et al, 2018;Springer et al, 2018). The need for non-homology-based functional annotations is pressing in maize, particularly as there is evidence these new and variably-present genes may be involved in hybrid vigor Baldauf et al, 2016Baldauf et al, , 2018. In this project, we evaluated the potential for using supervised machine learning-based classification algorithms to predict the function of annotated maize genes using purely non-homolog-based predictive variables.…”