Bulked Segregant RNA-seq Reveals Differential Expression and SNPs of Candidate Genes Associated with Waterlogging Tolerance in Maize

Abstract

Waterlogging has increasingly become one of the major constraints to maize productivity in some maize production zones because it causes serious yield loss. Bulked segregant RNA-seq (BSR-seq) has been widely applied to profile candidate genes and map associated Single Nucleotide Polymorphism (SNP) markers in many species. In this study, 10 waterlogging sensitive and eight tolerant inbred lines were selected from 60 maize inbred lines with waterlogging response determined and preselected by the International Maize and Wheat Improvement Center (CIMMYT) from over 400 tropical maize inbred lines. BSR-seq was performed to identify differentially expressed genes and SNPs associated with waterlogging tolerance. Upon waterlogging stress, 354 and 1094 genes were differentially expressed in the tolerant and sensitive pools, respectively, compared to untreated controls. When tolerant and sensitive pools were compared, 593 genes were differentially expressed under untreated and 431 genes under waterlogged conditions, of which 122 genes overlapped. To validate the BSR-seq results, the expression levels of six genes were determined by qRT-PCR. The qRT-PCR results were consistent with BSR-seq results. Comparison of allelic polymorphism in mRNA sequences between tolerant and sensitive pools revealed 165 (normal condition) and 128 (waterlogged condition) high-probability SNPs. We found 18 overlapping SNPs with genomic positions mapped. Eighteen SNPs were contained in 18 genes, and eight and nine of 18 genes were responsive to waterlogging stress in tolerant and sensitive lines, respectively. Six alleles of the 18 originated from tolerant pool were significantly up-regulated under waterlogging, but not those from sensitive pool. Importantly, one allele (GRMZM2G055704) of the six genes was mapped between umc1619 and umc1948 on chromosome 1 where a QTL associated with waterlogging tolerance was identified in a previous research, strongly indicating that GRMZM2G055704 is a candidate gene responsive to waterlogging. Our research contributes to the knowledge of the molecular mechanism for waterlogging tolerance in maize.

Keywords: RNA-seq; SNPs; abiotic stress; bulk segregant analysis; maize (Zea mays L.); waterlogging stress tolerance.

FIGURE 1

Differences in plant growth traits between tolerant and sensitive inbred groups upon waterlogging stress. (A) Relative shoot height, (B) relative shoot dry weight, (C) Survival rate. All experiments included three biological replications. Six germinated seeds were planted into each pot. For each inbred line, three pots containing 18 seedlings at the three-leaf stage were subjected to waterlogging treatment. After 8 days of waterlogging treatment, traits including survival rate and shoot height were measured. After measurement, shoots and roots were excised, transferred into envelope, and placed in an oven (65°C) for 3 days; shoot weight was then measured. Shoot height and shoot dry weight of waterlogged plants were compared to those of untreated plants. Each bar (in cm or g) represents the mean value ± SD of three independent analyses. One-way ANOVA test was performed to reveal the significance between tolerant and sensitive groups under waterlogging treatment. Among these lines, the different letters on the bars mean significant difference.

FIGURE 2

Changes of expression levels for six genes were revealed by RNA-seq and validated by qRT-PCR. (A–F) Relative expression level of GLK11, GRMZM2G132185, GRMZM2G148772, GRMZM2G086573, GRMZM2G022958, and GRMZM2G428554, respectively. Approximately 0.2 g of roots were excised from six seedlings and used for RNA isolation. Six genes were selected based on RNA-seq results, and their expression levels were validated by qRT-PCR. All experiments included three biological replications. Each bar represents the fold change in expression level of the gene by comparing waterlogged with normal conditions. Student’s t-test was performed. ^∗Indicates p < 0.05; ^∗∗indicates p < 0.01.

FIGURE 3

The GO term analysis of DEGs between tolerant and sensitive pools. GO term analysis of significant DEGs between tolerant and sensitive pools at normal condition (A) and under waterlogging stress (B). GO analysis was performed using GO Term Finder. The red and blue bars represent the numbers of up- and down-regulated genes between tolerant and sensitive pools, respectively. The GO terms are shown on the y-axis and the number of significant DEGs is at the end of each bar.

FIGURE 4

Putative SNPs associated with waterlogging stress. The polymorphic SNPs between tolerant and sensitive pools under normal condition (A) and waterlogging treatment (B). The polymorphic SNPs were identified by comparing RNA sequences of tolerant and sensitive pools. The linkage probability of each SNP marker associated with the causal gene was obtained from a Bayesian BSA analysis. The physical position of each SNP marker (x-axis) was plotted vs. the linkage probability (y-axis). The red solid line indicates the threshold of linkage probability (0.9).

FIGURE 5

Locations of 18 genes containing significant SNPs on the maize chromosomes. The scale represents a 20 Mb chromosomal distance. The numbers in distance are in the scale of Mb. Chromosome numbers are indicted at the bottom end of each chromosome.

FIGURE 6

The relative expression level of 18 genes with high-probability under waterlogging stress. (A) The expression level in tolerant line (CML495). (B) The expression level in sensitive line (CMTL001). Approximately 0.2 g of roots were excised from six seedlings and used for RNA isolation. A total of 18 genes with significant SNP overlapped between tolerant and sensitive pool under waterlogging and normal conditions, and their expression levels were validated by qRT-PCR. All experiments included three biological replications. Each bar represents the expression level of gene. Student’s t-test was performed. ^∗Indicates p < 0.05; ^∗∗indicates p < 0.01.