Genetic dissection of maize plant architecture with an ultra-high density bin map based on recombinant inbred lines

Abstract

Background: Plant architecture attributes, such as plant height, ear height, and internode number, have played an important role in the historical increases in grain yield, lodging resistance, and biomass in maize (Zea mays L.). Analyzing the genetic basis of variation in plant architecture using high density QTL mapping will be of benefit for the breeding of maize for many traits. However, the low density of molecular markers in existing genetic maps has limited the efficiency and accuracy of QTL mapping. Genotyping by sequencing (GBS) is an improved strategy for addressing a complex genome via next-generation sequencing technology. GBS has been a powerful tool for SNP discovery and high-density genetic map construction. The creation of ultra-high density genetic maps using large populations of advanced recombinant inbred lines (RILs) is an efficient way to identify QTL for complex agronomic traits.

Results: A set of 314 RILs derived from inbreds Ye478 and Qi319 were generated and subjected to GBS. A total of 137,699,000 reads with an average of 357,376 reads per individual RIL were generated, which is equivalent to approximately 0.07-fold coverage of the maize B73 RefGen_V3 genome for each individual RIL. A high-density genetic map was constructed using 4183 bin markers (100-Kb intervals with no recombination events). The total genetic distance covered by the linkage map was 1545.65 cM and the average distance between adjacent markers was 0.37 cM with a physical distance of about 0.51 Mb. Our results demonstrated a relatively high degree of collinearity between the genetic map and the B73 reference genome. The quality and accuracy of the bin map for QTL detection was verified by the mapping of a known gene, pericarp color 1 (P1), which controls the color of the cob, with a high LOD value of 80.78 on chromosome 1. Using this high-density bin map, 35 QTL affecting plant architecture, including 14 for plant height, 14 for ear height, and seven for internode number were detected across three environments. Interestingly, pQTL10, which influences all three of these traits, was stably detected in three environments on chromosome 10 within an interval of 14.6 Mb. Two MYB transcription factor genes, GRMZM2G325907 and GRMZM2G108892, which might regulate plant cell wall metabolism are the candidate genes for qPH10.

Conclusions: Here, an ultra-high density accurate linkage map for a set of maize RILs was constructed using a GBS strategy. This map will facilitate identification of genes and exploration of QTL for plant architecture in maize. It will also be helpful for further research into the mechanisms that control plant architecture while also providing a basis for marker-assisted selection.

Fig. 1

Genome-wide distribution of SNPs and genetic variants throughout the Ye478 and Qi319 genomes. The outermost box with scale represents the 10 maize chromosomes. The orange histogram represents the density of SNPs that are polymorphic between Ye478 and Qi319; the green histogram represents the density of polymorphic SNPs within coding sequences between Ye478 and Qi319; the blue histogram indicates the density of insertions or deletions (Indels) between Ye478 and Qi319

Fig. 2

Recombination bin map of the RIL population derived from Ye478 and Qi319. The bin map is comprised of 4183 bin markers inferred from 88,268 high-quality SNPs mapped in the RIL population. Physical position is based on B73 RefGen_V3 sequence. Red: Qi319 genotype; blue: Ye478 genotype; yellow: heterozygote

Fig. 3

Mapping of P1, which controls cob color, in the RIL population. Curves in plot indicate the genetic coordinates along chromosomes or the physical coordinates within a chromosome (x-axis) and LOD score (y-axis) of the detected QTL. Mapping curve of the QTL that controls cob color of is located on chromosome 1; the box shows a magnification of the peak on chromosome 1. The red dot represents the relative physical position of the P1 gene

Fig. 4

Variation in PH, EH, and IN was attributed to genetic and environmental factors across the RIL population. The different shades of grey in the stacked bar diagram indicate the various factors that explain phenotypic variance. PH: plant height; EH: ear height; IN: internode number

Fig. 5

Correlations between variation in PH, EH, and IN. Positive correlations between PH and EH were greater among line means than those with IN across all three environments. PH: plant height; EH: ear height; IN: internode number. Red arrow: Ye478; green arrow: Qi319; orange arrow: mid-parent. a, 2013 Shunyi; b, 2013 Gongzhuling; and c, 2014 Gongzhuling

Fig. 6

Mapping of QTL on ten chromosomes for PH, EH, and IN across three environments. The curves indicate the physical position (x-axis) of bin markers against LOD score (y-axis) of QTL detected on ten chromosomes. Different colors represent different environments: E1, 2013 Shunyi; E2, 2013 Gongzhuling; and E3, 2014 Gongzhuling. The red dashed lines present the LOD threshold. PH: plant height; EH: ear height; IN: internode number