A GBS-Based GWAS Analysis of Leaf and Stripe Rust Resistance in Diverse Pre-Breeding Germplasm of Bread Wheat (Triticum aestivum L.)

Abstract

Yellow (YR) and leaf (LR) rusts caused by Puccinia striiformis f. sp. tritici (Pst) and Puccinia triticina, respectively, are of utmost importance to wheat producers because of their qualitative and quantitative effect on yield. The search for new loci resistant to both rusts is an ongoing challenge faced by plant breeders and pathologists. Our investigation was conducted on a subset of 168 pre-breeding lines (PBLs) to identify the resistant germplasm against the prevalent local races of LR and YR under field conditions followed by its genetic mapping. Our analysis revealed a range of phenotypic responses towards both rusts. We identified 28 wheat lines with immune response and 85 resistant wheat genotypes against LR, whereas there were only eight immune and 52 resistant genotypes against YR. A GWAS (genome-wide association study) identified 190 marker-trait associations (MTAs), where 120 were specific to LR and 70 were specific to YR. These MTAs were confined to 86 quantitative trait loci (QTLs), where 50 QTLs carried MTAs associated with only LR, 29 QTLs carried MTAs associated with YR, and seven QTLs carried MTAs associated with both LR and YR. Possible candidate genes at the site of these QTLs are discussed. Overall, 70 PBLs carried all seven LR/YR QTLs. Furthermore, there were five PBLs with less than five scores for both LR and YR carrying positive alleles of all seven YR/LR QTLs, which are fit to be included in a breeding program for rust resistance induction.

Keywords: GWAS; breeding; leaf rust; resistance; wheat; yellow rust.

Figure 1

Overlaid histograms showing frequency distribution of (a) leaf rust disease severity (LRDS), (b) leaf rust co-efficient of infection (LRCI), (c) yellow rust disease severity (YRDS), and (d) yellow rust co-efficient of infection (YRCI) in season 1 (S1, green) and season 2 (S2, orange) where thin dashed colored lines indicate respective means. Tables below each histogram show mean ± S.E and range, whereas P indicates significant differences between genotypes (G), seasons (S), blocks (B), G × B interaction, G × S interaction, B × S interaction, and G × B × S interaction at, 0.01 (**) and 0.001 (***) significance level (p-value).

Figure 2

Correlation among disease severity (DS) and co-efficient of infection (CI) of leaf rust (LR) and yellow rust (YR) in season 1 (_S1) and season 2 (_S2). *, ** and *** indicated significance at 0.05, 0.01 and 0.001 p-value, respectively.

Figure 3

Genome-wide scan of (a) LRDS in S1 and (b) S2 in the form of Manhattan plots where the chromosomes are plotted at the bottom and the thin dotted red line indicates significance level at p-value < 0.001 (−log10 = 3 or more) beyond which an association is counted as true association (highlighted dots). Table on the right shows common MTAs between S1 and S2 showing marker names, chromosome number, position, p-value, and R² value in each season.

Figure 4

Genome-wide scan of (a) LRCI in S1 and (b) S2 in the form of Manhattan plots where the chromosomes are plotted at the bottom, and the thin dotted red line indicates significance level at p-value < 0.001 (−log10 = 3 or more) beyond which an association is counted as true association (highlighted dots). Table on the right shows common MTAs between S1 and S2, showing marker names, chromosome number, position, p-value, and R² value in each season.

Figure 5

Genome-wide scan of (a) YRDS in S1 and (b) S2 in the form of Manhattan plots where the chromosomes are plotted at the bottom, and the thin dotted red line indicates significance level at p-value < 0.001 (−log10 = 3 or more) beyond which an association is counted as true association (highlighted dots).

Figure 6

Genome-wide scan of (a) YRCI in S1 and (b) S2 in the form of Manhattan plots where the chromosomes are plotted at the bottom, and the thin dotted red line indicates significance level at p-value < 0.001 (−log10 = 3 or more) beyond which an association is counted as true association (highlighted dots). Table on the right shows common MTAs between S1 and S2, showing marker names, chromosome number, position, p-value, and R² value in each season.

Figure 7

Distribution of leaf rust (red lines), yellow rust (green lines), and leaf/yellow rust (blue lines in the inner circle) QTLs. Light-brown lines in the outer track indicate the marker positions on each chromosome; grey bars in the second circle indicate the number of markers confined to each QTL. The red, green, and/or blue lines under the track circle indicate the span of QTLs, with small vertical lines point to the peak position of QTL. See Table S3 for details.

Figure 8

Effect of increasing number of yellow and leaf rust (YR/LR) QTLs on all disease scores.