Prioritizing quantitative trait loci for root system architecture in tetraploid wheat

Abstract

Optimization of root system architecture (RSA) traits is an important objective for modern wheat breeding. Linkage and association mapping for RSA in two recombinant inbred line populations and one association mapping panel of 183 elite durum wheat (Triticum turgidum L. var. durum Desf.) accessions evaluated as seedlings grown on filter paper/polycarbonate screening plates revealed 20 clusters of quantitative trait loci (QTLs) for root length and number, as well as 30 QTLs for root growth angle (RGA). Divergent RGA phenotypes observed by seminal root screening were validated by root phenotyping of field-grown adult plants. QTLs were mapped on a high-density tetraploid consensus map based on transcript-associated Illumina 90K single nucleotide polymorphisms (SNPs) developed for bread and durum wheat, thus allowing for an accurate cross-referencing of RSA QTLs between durum and bread wheat. Among the main QTL clusters for root length and number highlighted in this study, 15 overlapped with QTLs for multiple RSA traits reported in bread wheat, while out of 30 QTLs for RGA, only six showed co-location with previously reported QTLs in wheat. Based on their relative additive effects/significance, allelic distribution in the association mapping panel, and co-location with QTLs for grain weight and grain yield, the RSA QTLs have been prioritized in terms of breeding value. Three major QTL clusters for root length and number (RSA_QTL_cluster_5#, RSA_QTL_cluster_6#, and RSA_QTL_cluster_12#) and nine RGA QTL clusters (QRGA.ubo-2A.1, QRGA.ubo-2A.3, QRGA.ubo-2B.2/2B.3, QRGA.ubo-4B.4, QRGA.ubo-6A.1, QRGA.ubo-6A.2, QRGA.ubo-7A.1, QRGA.ubo-7A.2, and QRGA.ubo-7B) appear particularly valuable for further characterization towards a possible implementation of breeding applications in marker-assisted selection and/or cloning of the causal genes underlying the QTLs.

Keywords: Association mapping; GWAS; drought stress; germplasm collection; grain yield; meta-QTLs; root growth angle; root system architecture; rooting depth; seminal root..

Fig. 1.

Measurement of root growth angle (RGA) in seminal roots at the seedling stage and in excavated roots of field-grown plants at the end of flowering. The two parental genotypes Colosseo and Lloyd, highly contrasted for RGA at both stages, are shown as an example. Direct measurements were carried out for seminal root RGA, while software-aided digital measurements were carried out for adult root systems (REST software). (This figure is available in colour at JXB online.)

Fig. 2.

Root system architecture (RSA) of seedlings and field-grown plants of durum wheat parental lines and accessions from the elite Unibo-DP showing contrasting phenotypes for RSA features. (A) Comparison of RSA for the parental lines Colosseo (left) and Lloyd (right) at the adult stage in the field. Dotted lines originating at the crowns were traced to delimit 95% of the 2-D projected area of the root system (as suggested in the REST software manual). (B) Comparison of seedling and adult RGA for the five Unibo-DP accessions with the narrowest RGA according to seedling measurements. (C) Comparison of seedling and field RGA for the five Unibo-DP accessions with the widest RGA according to seedling measurements. (This figure is available in colour at JXB online.)

Fig. 3.

Relationship between seminal root growth angle (RGA) at the seedling stage and adventitious root system growth in the adult plant under field conditions for a selection of 44 Unibo-DP accessions. Three accession groups were considered: (i) 12 Unibo-DP accessions bottom-ranking for seminal RGA; (ii) 12 Unibo-DP accessions top-ranking for seminal RGA; and (iii) 16 parents of mapping populations developed at UNIBO plus the parental lines of the Colosseo×Lloyd and Meridiano×Claudio RIL populations.

Fig. 4.

Genetic maps of root system architecture (RSA) QTL clusters and root growth angle (RGA) QTLs for chromosomes 2B, 4B, 6A, and 7A. The reference map is the tetraploid consensus map reported by Maccaferri et al. (2015). RSA and RGA QTLs from Co×Ld and Mr×Cl RIL populations, GWAS QTLs from Unibo-DP. and previously published QTLs from bread wheat studies have been projected onto the reference maps. Single-component QTLs are reported as vertical bars corresponding to confidence intervals. RSA QTL clusters are highlighted by horizontal shaded banding. GY and TGW QTLs in the trait acronyms are the same as in Table 1. QTL significance levels are highlighted using ** for suggestive QTLs, *** for nominal QTLs, and **** for GWAS experiment-wise significant QTLs (Unibo-DP only).

Fig. 4.

Genetic maps of root system architecture (RSA) QTL clusters and root growth angle (RGA) QTLs for chromosomes 2B, 4B, 6A, and 7A. The reference map is the tetraploid consensus map reported by Maccaferri et al. (2015). RSA and RGA QTLs from Co×Ld and Mr×Cl RIL populations, GWAS QTLs from Unibo-DP. and previously published QTLs from bread wheat studies have been projected onto the reference maps. Single-component QTLs are reported as vertical bars corresponding to confidence intervals. RSA QTL clusters are highlighted by horizontal shaded banding. GY and TGW QTLs in the trait acronyms are the same as in Table 1. QTL significance levels are highlighted using ** for suggestive QTLs, *** for nominal QTLs, and **** for GWAS experiment-wise significant QTLs (Unibo-DP only).

Fig. 4.

Genetic maps of root system architecture (RSA) QTL clusters and root growth angle (RGA) QTLs for chromosomes 2B, 4B, 6A, and 7A. The reference map is the tetraploid consensus map reported by Maccaferri et al. (2015). RSA and RGA QTLs from Co×Ld and Mr×Cl RIL populations, GWAS QTLs from Unibo-DP. and previously published QTLs from bread wheat studies have been projected onto the reference maps. Single-component QTLs are reported as vertical bars corresponding to confidence intervals. RSA QTL clusters are highlighted by horizontal shaded banding. GY and TGW QTLs in the trait acronyms are the same as in Table 1. QTL significance levels are highlighted using ** for suggestive QTLs, *** for nominal QTLs, and **** for GWAS experiment-wise significant QTLs (Unibo-DP only).