Mapping of adult plant leaf rust resistance in Aus27506 and validation of underlying loci by in-planta fungal biomass accumulation

Among the rust diseases, leaf rust of wheat caused by Puccinia triticina , is the most prevalent worldwide and causes significant yield losses. This study aimed to determine the genomic location of loci that control adult plant resistance (APR) to leaf rust in the pre-Green Revolution landrace accession, Aus27506, from the ‘Watkins Collection’. An Aus27506/Aus27229-derived F7 recombinant inbred (RIL) population was screened under field conditions across three cropping seasons and genotyped with the iSelect 90K Infinium SNP bead chip array. One QTL on each of chromosomes 1BL, 2B and 2DL explained most of the leaf rust response variation in the RIL population and were named QLr.sun-1BL , QLr.sun-2B and QLr.sun-2DL , respectively. QLr.sun-1BL and QLr.sun-2DL were contributed by Aus27506. QLr.sun-1BL is likely Lr46 , while QLr.sun-2DL appeared to be a new APR locus. The alternate parent, Aus27229, carried the putatively new APR locus QLr.sun-2B. Comparisons of average severities among RILs carrying these QTL in different combinations indicated that QLr.sun-2B does not interact with either of the other two QTL; however, the combination of QLr.sun-1BL and QLr.sun-2DL reduced disease severity significantly. In-planta fungal quantification assays validated these results. The RILs carrying QLr.sun-1BL and QLr.sun-2DL did not differ significantly from parent Aus27506 in resistance. Aus27506 can be used as a source of adult plant leaf rust resistance in breeding programs. of Lr34. SNPLr46G22 the same allele in both parents suggesting the presence of Lr46. enhanced In-plant fungal quantification using chitin assay the additive effect of resistance loci RILs carrying two QTL combinations involving QLr.sun-2B showed more fungal biomass compared to the combination of QLr.sun-1BL and QLr.sun-2D. These results confirm conclusions drawn from field disease severity score comparison of RILs possessing different combinations of QTL.

Transfer of genetically diverse resistance genes in wheat cultivars is the most costeffective way to control rust diseases [5]. Leaf rust resistance genes can be divided into two classes based on the plant growth stage at which resistance is expressed [7]. Most leaf rust resistance genes are all stage resistance (ASR) genes that are effective against avirulent races throughout plant growth [4]. ASR genes often confer high levels of resistance, but they can be rapidly defeated by pathogen evolution. In contrast, the second class of genes, adult plant resistance (APR) genes, only provide resistance in mature plants. APR is typically only partially effective and not associated with hypersensitive host cell death [5]. However, combinations of two or more APR genes can provide commercially acceptable or near immune levels of resistance and this type of resistance is assumed to be durable [6,38]. However, a few atypical APR genes also exist, for example Lr22b is race-specific and expresses high levels of hypersensitive resistance at adult plant stages suggesting that mechanistically it is more similar to ASR genes.
Over the last two decades, rust resistance breeding has reduced the deployment of ASR in favor of APRs and/or combinations of both types through marker-assisted selection [4].
APR expression under field conditions can be detected by different methods, such as area under disease progress [17], size and number of uredinia produced during disease development and latent period [20,36]. These measurements are laborious and require specialized skills. To overcome the difficulties of these time-consuming disease assessment methods Ayliffe et al.
Eighty QTL for leaf rust resistance have been mapped [23] and many have been detected in multiple studies.. Both wheat landraces and close relative species of wheat have been used to discover new sources of disease resistance [2,11]. In this study a pre-Green Revolution wheat landrace collected from France (Aus27506), and is susceptible to leaf rust at the seedling stage but has adult plant resistance, was genetically dissected to identify QTL underlying the leaf rust APR. Three QTL were identified, and their relative effectiveness and additivity was measured using chitin based in-planta fungal quantification.

Development of mapping population
Leaf rust resistant landrace Aus27506 was selected from the 'Watkins Collection' [2] and crossed with the moderately susceptible landrace, Aus27229. A recombinant inbred line (RIL) population consisting of 106 RILs (F6:7) was developed.

Field evaluation
Eight to ten seeds of each RIL and both parents were sown as hill-plots at the experimental Adult plant leaf rust responses were scored from flag leaf initiation to grain filling at a weekly interval based on a 1-9 scale [7]. The 1-9 scale was converted to a disease severity score [10] to allow RILs carrying different combinations of QTL for leaf rust resistance to be compared.

DNA isolation and quantification
Genomic DNA was extracted from the RILs and parents using a modified CTAB method [3] and quantified using a Nanodrop ND-1000 spectrophotometer (Nanodrop Technologies).
iSelect 90K Infinium bead chip array genotyping The RIL population was genotyped at AgriBio, La Trobe University, Melbourne, using the iSelect 90K Infinium SNP bead chip array described by Wang et al. [43].
RILs that did not produce high quality genotypic data (≥20 % missing data) were excluded from downstream analyses. Monomorphic markers and markers with more than 10 % data missing was also excluded. The remaining marker data was evaluated using Chi-squared analysis and markers that deviated from a 1:1 segregation ratio were discarded (χ²(1:1) = 3.94, non-significant at P = 0.05). Markers with 5% or less heterozygous calls were retained to avoid false purging of heterozygous loci [9].

Linkage map construction and QTL analysis
MapManager version QTXb20 [25] was used for genetic linkage map construction. The Kosambi mapping function [19] was used to convert recombination fractions into centiMorgans (cM). Redundant markers were excluded using the command 'hide redundant loci' option and phenotypic data were imported into MapManager. QTL cartographer [42] was used for composite interval mapping (CIM) based on 1000 permutations.

Genotyping with markers linked with known APR genes
To screen the parental lines for the presence of previously characterised leaf rust APR genes the following markers were tested; csLV34 for the Lr34 gene [22], SNPLr46G22 for the Lr46 gene (Lagudah unpublished) and csGS, cs7BLNLRR and Psy1-1 for the Lr68 gene [16]. These markers were amplified using standard PCR conditions except for KASP marker SNPLr46G22 which used the KASP assay described in Chhetri et al. [10]. Markers revealing polymorphism between the parents were tested on the entire RIL population and incorporated into the genetic map.

Statistical analysis
Chi squared analysis was used to test the goodness of fit of the observed segregation to the expected genetic ratios. Wright's formula [45]  Fungal chitin was quantified in these samples using the method described by Ayliffe et al. [1].
Briefly each sample was autoclaved at 121°C and 15 psi for 20 minutes with loosened caps. The

Genotyping with markers linked with known genes
Genotyping of the parental lines with marker csLV34 indicated the absence of Lr34. Marker SNPLr46G22 produced the same allele in both parents suggesting the presence of Lr46.

Markers csGS, cs7BLNLRR and Psy1-1 were monomorphic between parents and produced
Lr68 specific amplicon indicating the presence of Lr68 in both parents.

Quantification of fungal biomass by chitin assay
To assess the fungal biomass, infected flag leaves of the parents and representative RILs carrying different QTL combinations were collected and used for the chitin assay. Aus27506 and Aus27229 differed significantly for fungal biomass (Fig. 3). Fungal growth in parent Aus27506, which carried QLr.sun-2DL and QLr.sun-1BL, was 61% lower than that of Aus27229 which carried QLr.sun-2B. The RILs carrying all three QTL did not differ significantly in fungal biomass accumulation compared with Aus27506. RILs with QLr.sun-2DL + QLr.sun-1BL had more fungal colonization compared with Aus27506 and RILs with all three QTL but significantly reduced fungal growth compared to RILs with the other two dual gene combination (QLr.sun-1BL+QLr.sun-2B and QLr.sun-2B+QLr.sun-2DL) (Fig. 3). The dual combinations involving QLr.sun-2B, contributed by Aus27229 and either of QLr.sun-1BL or QLr.sun-2DL showed similar levels of fungal growth which was less than that of Aus27229 (QLr.sun-2B).

Composite interval mapping of adult plant leaf rust response variation among an
Aus27506/Aus27229 RIL population identified three QTL on chromosomes 1BL, 2B and 2DL, respectively. Chromosome 1B carries formally designated leaf rust resistance genes Lr26/Yr9/Sr31 [24], Lr33 [12], Lr44 [13], Lr46 [39], Lr51 [15], Lr55 [28], Lr71 [37] and Lr75 [40]. Of these genes, only Lr33 and Lr46, are located on the long arm. Lr33 is located 3 cM distal to the centromere [12] and Lr46 is located in the most distal deletion bin (1BL-0.84-0.89) of the long arm [26]. QLr.sun-1BL detected in this study was located in the distal region of chromosome 1BL (137 cM of a total map length of 174 cM). Screening with the Lr46-linked marker SNPLr46G22 did not differentiate the parents presumably due to the non-diagnostic nature of this marker and consequently false positive amplification in Aus27229. QLr.sun-1BL explained (11-22 %) of the phenotypic variation which is similar to that reported for Lr46 in other studies [23]. Further Lr46 appears to express better in cooler climates [21]. Many studies have similarly reported disease severity to be lowered by combinations of APR genes; for example, Lr34 and Lr46, Lr34 and Lr68, Lr75 and QLr.sfr-7BL showed enhanced resistance through additive gene action in different trials [40]. In-plant fungal quantification using chitin assay is another measure to show the additive effect of resistance loci [1]. RILs carrying two QTL combinations involving QLr.sun-2B showed more fungal biomass compared to the combination of QLr.sun-1BL and QLr.sun-2D. These results confirm conclusions drawn from field disease severity score comparison of RILs possessing different combinations of QTL.