Genome-wide association study (GWAS) reveals genetic loci of lead (Pb) tolerance during seedling establishment in rapeseed (Brassica napus L.)

Abstract

Background: Lead (Pb) pollution in soil has become one of the major environmental threats to plant growth and human health. Safe utilization of Pb contaminated soil by phytoremediation require Pb-tolerant rapeseed (Brassica napus L.) accessions. However, breeding of new B. napus cultivars tolerance to Pb stress has been restricted by limited knowledge on molecular mechanisms involved in Pb tolerance. This work was carried out to identify genetic loci related to Pb tolerance during seedling establishment in rapeseed.

Results: Pb tolerance, which was assessed by quantifying radicle length (RL) under 0 or 100 mg/L Pb stress condition, shown an extensive variation in 472 worldwide-collected rapeseed accessions. Based on the criterion of relative RL > 80%, six Pb-tolerant genotypes were selected. Four quantitative trait loci (QTLs) associated with Pb tolerance were identified by Genome-wide association study. The expression level of nine promising candidate genes, including GSTUs, BCATs, UBP13, TBR and HIPP01, located in these four QTL regions, were significantly higher or induced by Pb in Pb-tolerant accessions in comparison to Pb-sensitive accessions.

Conclusion: To our knowledge, this is the first study on Pb-tolerant germplasms and genomic loci in B. napus. The findings can provide valuable genetic resources for the breeding of Pb-tolerant B. napus cultivars and understanding of Pb tolerance mechanism in Brassica species.

Keywords: GWAS; Lead (Pb) tolerance; Phytoremediation; Rapeseed; SNP markers.

Fig. 1

Distributions and correlation matrixes of traits. a Violin plot of radicle length (RL) under control (CK) and Pb stress (Pb) condition. b Distributions and correlation matrixes of relative radicle length (RRL). RRL1, RRL2, RRL3 represent the RRL in replication 1, 2 and 3 respectively. RRL_Means was the average value of three RRLs

Fig. 2

Manhattan plots of association analysis for RRLs using Q + K model. The red, pink, blue and green dots represent the association signals for RRL_Means (average value of three RRLs), RRL1 (RRL in replication 1), RRL2 (RRL in replication 2) and RRL3 (RRL in replication 3), respectively. The blue and red horizontal lines indicate the significantly associated threshold (−log10(1/19,945) = 4.3) and moderately associated threshold (−log10(p) between 3.5–4.3), respectively

Fig. 3

Allele effects of associated SNPs. Red, green, blue and purple boxes indicated the A, C, G and T alleles, respectively. The “P” presents significant different level of RRL between alleles by t- test

Fig. 4

Association mapping for RRL on chromosome C03. Plots show the SNPs in the QTL Pb-C03–1 (top left of Fig. 4, from 1,241,778 to 1,446,328 bp on chromosome C03) and Pb-C03–2 (top right of Fig. 4, from 58,079,114 to 58,097,249 bp on chromosome C03) regions associated with RRL. The red, pink, blue and green dots represent the association signals for RRL_Means (average value of three RRLs), RRL1 (RRL in replication 1), RRL2 (RRL in replication 2) and RRL3 (RRL in replication 3), respectively. The blue and red horizontal line indicate the threshold of significantly associated SNPs at −log10 (1/19,945) = 4.3 and threshold of moderately associated SNPs at 3.5 ≤ −log10 (p) ≤ 4.3, respectively as in Fig. 2. The heat maps span the linkage disquilibrium (LD) region with the most strongly associated SNPs (r² > 0.4)

Fig. 5

The relative expression level of candidate genes for Pb tolerance. The “a, b, c, d and e” present significant difference at 5% level by tukey test