Phenotypic and genome-wide association analyses for nitrogen use efficiency related traits in maize (Zea mays L.) exotic introgression lines

Abstract

Nitrogen (N) limits crop production, yet more than half of N fertilizer inputs are lost to the environment. Developing maize hybrids with improved N use efficiency can help minimize N losses and in turn reduce adverse ecological, economical, and health consequences. This study aimed to identify single nucleotide polymorphisms (SNPs) associated with agronomic traits (plant height, grain yield, and anthesis to silking interval) under high and low N conditions. A genome-wide association study (GWAS) was conducted using 181 doubled haploid (DH) lines derived from crosses between landraces from the Germplasm Enhancement of Maize (BGEM lines) project and two inbreds, PHB47 and PHZ51. These DH lines were genotyped using 62,077 SNP markers. The same lines from the per se trials were used as parental lines for the testcross field trials. Plant height, anthesis to silking interval, and grain yield were collected from high and low N conditions in three environments for both per se and testcross trials. We used three GWAS models, namely, general linear model (GLM), mixed linear model (MLM), and Fixed and Random model Circulating Probability Unification (FarmCPU) model. We observed significant genetic variation among the DH lines and their derived testcrosses. Interestingly, some testcrosses of exotic introgression lines were superior under high and low N conditions compared to the check hybrid, PHB47/PHZ51. We detected multiple SNPs associated with agronomic traits under high and low N, some of which co-localized with gene models associated with stress response and N metabolism. The BGEM panel is, thus, a promising source of allelic diversity for genes controlling agronomic traits under different N conditions.

Advancements and Prospects of Genome-Wide Association Studies (GWAS) in Maize

“…Additionally, two independent methods were developed: PEPIS for multiple hybrid populations (MHPs) corresponded with compressed mixed linear model (CMLM) to discover candidate genes and QTNs for maize in GWAS [60]. Moreover, GWAS was conducted using a maize heterosis population for genetic variation among populations, and three models were implemented, namely, GLM, MLM, and FarmCPU [61], and an additive genetic model was used for per se trials, whereas dominant models were used for test cross trials. Nevertheless, additional algorithms in the R program GAPIT were used to performed GWAS for each model [62][63][64].…”