Spot form of net blotch is an important barley disease, and along with net form net blotch, it can cause serious damage to barley production in the absence of appropriate management measures. Although the disease can be controlled with fungicides and good agricultural practices, host resistance remains the most economically and environmentally friendly option to manage Ptm. Several major and minor genes associated with SFNB resistance have been identified in previous studies using bi-parental mapping (reviewed by Liu et al. 2011) and recent GWAS approaches (Wang et al. 2015; Tamang et al. 2015; Burlakoti et al. 2017). Because SFNB undergoes frequent sexual recombination, there is a high risk of increased virulence within the pathogenic populations due to natural selection. Thus, identifying new sources of resistance is important for developing resistant cultivars which recombine multiple resistance loci. In this study, we deployed GWAS approach using mixed linear models accounting for population structure and relatedness to reduce false positives. We also used FDR correction of significant markers to minimize false positives and to detect true marker-trait associations (Yu et al. 2006; Zhao et al. 2007; Patterson et al. 2016).
The phenotypic variation reported in this study showed a high variability in SFNB resistance among barley genotypes at the seeding stage and the adult plant stage in diverse environments. Among the 336 barley genotypes, nine were highly resistant in at least five environments and exhibited an IRs ≤1.5 to Ptm isolates. These lines could be used as potential resistance sources for SFNB resistance in barley breeding programs. In this study, we found 42 loci associated with resistance/susceptibility on all seven barley chromosomes at both growth stages. Of the identified QTL, 11 correspond to previously known QTL involved in resistance/susceptibility to SFNB and/or net blotch resistance in general. These results validate the GWAS approach used in the current study. The remaining 31 QTL were not reported before elsewhere, therefore they were considered as novel and may be useful resources for developing SFNB resistant cultivars. However, majority of the QTL mapped in our study were specific to a given environment and the variability of the infection responses of the genotypes, which indicates the quantitative nature of SFNB resistance in our association mapping panel. This is not surprising, since quantitative resistance of SFNB infection was also reported in previous studies (Williams et al. 1999; Friesen et al. 2006; Wang et al. 2015; Tamang et al. 2015; Burlakoti et al. 2017). This can be explained by a minor-gene for minor-gene interaction, where minor effect virulence genes in the pathogen correspond to resistance genes of minor effect in the host, due the specificity of the pathogen isolates or races used for screening (Poland 2009). The QTL QRptm6H-1s which conferred resistance against a Moroccan Ptm isolate SM4-2, was also reported previously by Tamang et al. 2015 which conferred resistance against isolates FGO-Ptm (USA), NZKF2 (New Zealand) and DEN (Denmark). Our results suggest that SM4-2, FGO-Ptm, and NZKF2 may share common virulence genes though these isolates were originated from Morocco, USA and the New Zealand, respectively.
One QTL, QRptm-1H-1a, associated with adult-plant stage resistance was identified on chromosome 1H (4.11 cM) and explained 3.62% of the phenotypic variation. The same QTL (QRptta-1H-4.11) was found to be associated with net form net blotch in a previous association mapping study, suggesting that this locus may be linked with resistance/susceptibility to both forms of net blotch (Amezrou et al. 2018). On chromosome 2H, QRptm-2H-1a having the largest allelic effect is ~2cM away from SFNB-2H-38.03, a SFNB resistant QTL mapped by Burlakoti et al. (2017), from a combined population of four barley breeding programs in the Upper Midwest of the USA when challenged with SFNB isolates collected from Montana, USA. Similarly, QRptm-2H-4a, is located at ~2cM and ~5cM distance, respectively, from the QTLs mapped by Cakir et al. (2011) in a Baudin/AC Metcalfe DH population using NB320 isolate and a diverse sample of the BCC (Barley core collection), screened with the isolate FGO-Ptm (Tamang et al. 2015). The SNP markers SCRI_RS_170162 and SCRI_RS_157097 (QRptma-2H-2a) are predicted to encode an unknown protein and LRR receptor protein kinase, respectively. These two markers explained 3.41 and 3.64% of the phenotypic variation, respectively. This QTL (QRptma-2H-2a) was not previously reported, therefore the predicted genes containing these SNPs can be considered as candidate SNFB resistance/susceptibility genes. Using the New Zealand isolate NZKF2, Tamang et al. (2015) detected two marker-trait associations on chromosome 3H at 43.52 cM and 103 cM. This is likely to be same QTL as the QRptma-3H-1a and QRptma-3H-2a identified in this study, suggesting that these two QTL may be the same and confer resistance at both developmental stages.
Three marker-trait associations were found significant in chromosome 4H. The QTL QRptma-4H-1a falls in the range of QRptms4, a QTL mapped by Grewal et al. (2012) using the CDC Bold/TR251 double haploid population and the SFNB isolate WRS857. Similarly, QRptms-4H-1s resides within the net blotch resistance locus QRpts4, which explained up to 21% of the phenotypic variation (Grewal et al. 2008). The remaining two QTLs (QRptm-4H-2s and QRptma-4H-2a) were not previously reported to be associated with SFNB resistance/susceptibility. Interestingly, QRptma-4H-2a was also associated with resistance to net form net blotch and may represent a potential source of resistance to two closely but distinct pathogens (Amezrou et al. 2018). The QTL QRptm-5H-1 was detected in two environments and in SM30-1 isolate with five significant SNPs, indicating that these markers are linked together and co-segregate for SFNB resistance. This QTL may be same as Rpt6, a major SFNB resistance gene located at about 38 cM on 5H (Manninen et al. 2006). The remaining two QTL mapped on 5H (Table 3, 4) were not previously reported and therefore are consider novel.
The centromere region of chromosome 6H has long been associated with both net form and spot form of net blotch resistance/susceptibility. We identified a long-range genomic region in this specific region of 6H (60.71-86.97cM) associated with resistance in both growth stages that support previous findings. The SNP SCRI_RS_199940 located at 2.62 cM on 6HS explained 3.39% of phenotypic variation. Burlakoti et al. (2017) also reported a QTL on the same genomic region (SFNB-6H-5.4). Similarly, QRptma-6H-2a and QRptma-6H-4a were previously reported at the seedling stage by Tamang et al. (2015) and at both growth stages by Wang et al. (2015), respectively.
The Rpt4 gene on chromosome 7H was the first SFNB resistance gene described in the cultivar Galleon and flanked by the RFLP markers Xpsr117D and Xcdo673 at approximately 6 to 25 cM (Williams et al. 1999). We identified one locus at the Rpt4 region (QRptma-7H-2a) that explained 5.50% of the phenotypic variation at the adult-plant stage. Further, the most consistent QTL, QRptma-7H-3, detected in our study at both growth stages had 16 significant marker-trait associations at ~70-71 cM. This indicates that this underlying region is likely a cluster of SFNB resistance or susceptibility genes. Wang et al. (2015) found that the QTL with the largest effects were located on chromosome 7H. Our findings also suggest that 7H harbors several alleles of resistance and should be accumulated to breed high-level SFNB resistant cultivars.
In conclusion, we have detected most of the major and minor SFNB resistance QTL previously reported on chromosome 2H, 3H, 4H, 6H and 7H (Williams et al. 1999; Williams et al. 2003; Cakir et al. 2011; Grewal et al. 2012; Tamang et al. 2015; Wang et al. 2015; Burlakoti et al. 2017) validating our approach while we also reported new QTL on all seven barley chromosomes in this study. The loci identified in this study could harbor either resistance or susceptibility targets. Breeding strategies must combine multiple loci, either by eliminating host susceptibility targets or introgressing resistance loci. It is therefore important to dissect host-pathogen interactions and the genes/loci conditioning lack of susceptibility and resistance for an effective deployment of SFNB resistant genotypes. Also, the marker haplotype analysis of the significant SNP at each QTL of highly resistant barley lines should provide a useful resource for marker-assisted selection. These results provide important genetic information for an effective deployment of resistance or elimination of host susceptibility factors from elite barley genotypes and provides a durable means of management for this important barley disease.