Leaf rust disease caused by the fungal pathogen Puccinia triticina syn. P. recondita Rob. Ex. Desm. f.sp. tritici Eriks. & E. Henn is a significant threat to the yield of wheat crop in all major wheat growing parts of the world. Reports of yield loss in wheat due to damage by leaf rust range from 30-50% [1]. Plant breeders utilize the model of transferring leaf rust resistance genes (Lr genes) into the host in order to confer it with genetic resistance. However, the pathogen has been able to throw up physiological races to cause virulence against the deployedLr genes and convert the resistant variety into a susceptible one. Since it is expected thatLr genes sourced from wild relatives are likely to be more durable, several have been transferred into wheat from its wild relatives and many of these have been documented as located on different chromosomes [2,3]. The geneLr28 is one such gene transferred from Aegilops speltoides, which is assigned into bread wheat through a chromosomal translocation T4AS.4AL-7S #2S located on chromosome 4AL [2].Lr28 is an effective gene for resistance from seedling stage through the entire lifespan of wheat crop in most parts of the world including the South Asian wheat regions [4]. There are more than 60Lr genes available with varying degrees of resistance of which many are indistinguishable from each other in their phenotypic expression. Molecular markers serve the purpose by detecting only those plants that carry the distinct genes. In breeding populations, the phenotypic expression of resistance would be identical in plants which are either heterozygous or homozygous at the resistance locus but distinction between these categories is essential since the latter only are desirable to be carried forward. Dominant molecular markers such as RAPD, SCAR or AFLP markers also do not serve that purpose. The currently availableLr28 linked markers are only dominant type markers [5]. Though reported a null allelic SSR marker; it cannot be useful for direct selection. Such a marker could only be used for confirmation or zygosity determination in those plants which are already identified asLr28 positive through phenotyping or marker assisted selection utilizing other dominant markers. It has been already proved by that the codominant STS marker reported by was actually not associated withLr2 [6-8]. Pyramiding resistance genes in combination is an effective way of thwarting the breakdown of resistance and in providing diversity that limits race evolution. The current investigation to identify a codominant SSR marker polymorphic forLr28 gene locus employs one F₂ breeding population targeted at combining APR geneLr48 with the seedling resistance geneLr28. It is anticipated that combinations of effective seedling resistance genes with race non-specific APR genes may provide a longer lasting resistance [9].

The codominant SSR marker, Xwmc497 which is being reported in this paper as linked toLr28 locus was used to select plants which carried homozygousLr28 resistance alleles. The two dominant flanking RAPD markers, S3₄₅₀ linked to the recessive resistance allele and S336₇₇₅ linked to the dominant susceptibility allele at theLr48 locus, which span a distance of 11.3 cM were employed to identify the plants carryingLr48 recessive resistant allele alone [10]. Wheat genotypes from diverse genetic backgrounds which have been testified to carry various other alien and native genes were included in the study for validating the marker for Lr28.

Plant material

An F₂ population developed from the cross between the most widely cultivated and successful Indian wheat cultivar PBW343 carrying the gene Lr28 (PBW343-Lr28) developed at IARI, India and the Australian cultivar Condor derived CSP44 line (with WW80/2*WW1511Kalyansona parentage) carrying the geneLr48 (CSP44-Lr48) was used for the study.Lr28 is a seedling resistance gene thus conferring resistance in all stages of the plant andLr48 is an adult plant resistance gene, effective only from the time the plant reaches booting stage. The zygosity of each of the F₂ individual plants was established both by F₃ progeny testing and co-dominant molecular marker analysis. A set of 30 plants per each F₂ family were sown to erect the F₃ population. The experiments were conducted in the controlled conditions of National Phytotron Facility, IARI and New Delhi

Pathotype of the fungal pathogen

The inoculum of the most virulent Puccinia recondita pathotype, 77-5 (121R63-1) was obtained from the Directorate of Wheat Research, Regional Station, Flowerdale, Shimla. Inoculation of the spores of the pathotype was done by spraying inoculum suspended in water fortified with Tween-20® (0.75 μl/ml) at an average concentration of 20 urediospores/microscopic field (10x × 10x).

DNA extraction

Young leaves from parents and individuals of the segregating population were collected, lyophilized and ground in liquid nitrogen using a pestle and mortar. DNA extraction was performed by the microextraction method described by Prabhu et al. [11]. Final concentration of DNA samples was maintained at 10 μg/μl for PCR reactions.

Seedling test

After sampling for DNA extraction, seedlings 8-10 days old at decimal code DC 11 stage were inoculated during the evening hours [12]. Prior to inoculation, the plants were sprayed with water to provide a uniform layer of moisture on the leaf surface. After inoculation, the seedlings were incubated for 36 h in humid glass chambers at a temperature of 23 ± 2°C and more than 85% relative humidity after which, the pots were shifted to muslin cloth chambers in the same green house. The disease reaction was recorded 12-14 days after inoculation, using the scoring method described by Stakman et al. [13].

PCR Amplification using molecular markers

Ten SSR markers specific to the 4A chromosome were selected from published data [14,15]. The SSR markers (custom synthesized at Biobasic Inc, Canada) were used to screen the parents (PBW343-Lr28 and CSP44-Lr48), F₂ population (comprising homozygous resistant, homozygous susceptible and heterozygous plants) and bulks (resistant and susceptible).

PCR amplification was done following the protocol developed by Williams et al. [16]. The PCR reactions with SSR markers were performed in a 20 μl volume which consisted of 10 mM Tris HCl (pH 8.3), 50 mM KCl, 2 mM MgCl₂, 200 μM of each dNTP (MBI Fermentas, Germany), 40 ng of each of the forward and reverse primers, 0.75 U Taq DNA polymerase (Banglore Genei Pvt. Ltd., India) and 50 ng template DNA. PCR amplifications for RAPD markers were performed in 20 μl reaction volume containing 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 2 mM MgCl₂, 200 μM of each dNTP (MBI Fermentas, Germany), 0.2 μM of primer, 0.75 U Taq DNA Polymerase (Bangalore Genei Pvt. Ltd., India) and 10-15 ng of genomic DNA. The amplification reactions were carried in a PTC-200 thermal cycler (MJ Research, Las Vegas, NV, USA) with the following thermal profile – initial denaturation of 94°C for 10 min followed by 44 cycles of 94°C for 1 min (denaturation), 61°C and 36°C (for SSR markers and RAPD markers respectively) for 1 min (annealing), 72°C (extension) and a final extension step of 72°C for 10 min. This was followed by 4°C for 10 min.

The amplified products from SSR markers and RAPD markers were separated on a 3% Metaphor® agarose gel and 2% Agarose gel respectively, in 1X TAE buffer at 80 V for 3 hrs to separate the fragments. The gels were later stained with 10 mg/ml ethidium bromide and viewed in a digital gel documentation system (Alpha Innotech, San Leandro, CA, USA).

Bulked segregant analyses were done to identify the markers’ linkage to the dominant resistance gene [17]. Ten randomly selected plants from the homozygous resistant and homozygous susceptible F₂ plants were used to prepare bulks. The bulks differentiated for the presence and absence of the leaf rust resistance geneLr28 (Figure 1).

Figure 1: Screening of the SSR marker Xwmc497₂₉₁ on the bulked DNA constituent F₂ plants of the cross PBW343 X CSP44 for genetic linkage analysis. M: 100bp DNA ladder, Lanes1-10: F₂ seedling resistant individual plants, 11: Resistant Bulk, 12: Resistant parent, PBW343+ Lr28, 13-22: F₂ seedling susceptible individual plants, 23: Susceptible Bulk 24: Seedling susceptible parent, CSP44+ Lr48.

Statistical Analysis

Segregation ratios were analyzed using a chi-square test. The individuals from the crosses that were scored as resistant and susceptible in the progeny populations were subjected to chi-square test for goodness of fit to test the deviation from the theoretically expected Mendelian segregation ratios. Mean and standard error of the grain yield of the F₂ plants was calculated on the basis of standard formulae. The linkage analysis was carried out using Mapmaker version 3.0 [18].

The parent PBW343-Lr28 showed resistance to the 77-5 (121R63- 1) race of Puccinia triticina with a resistant infection type of 0; while the APR parent, CSP44 showed a typical seedling susceptibility with a reaction type of 33+ (Growth stage 11 of Zadoks growth scale). 61 seedlings of the F₂ population showed susceptibility to the leaf rust infection while the remaining 193 plants remained resistant by expressing the seedling resistance conferred by the dominant resistance allele of theLr28 locus and the population followed a monogenic segregation ratio (P = 0.6645). All the susceptible F₂ derived F₃ families remained susceptible whereas only 67 out of the 193 resistant F₂ derived F₃ families were homozygous for resistance. The remaining 126 families were heterozygous thus distributing the F₂ genotypes into 1R:2R:1S monogenic segregation ratio (P = 0.6467). The phenotypic expression of adult plant resistance could not be examined due to the interference of the dominant seedling resistance geneLr28 in the same genetic background.

Out of ten SSR markers specific to the 4AL chromosome, only Xwmc497 (Forward: 5’CCCGTGGTTTTCTTTCCTTCT3’, Reverse: 5’AACGACAGGGATGAAAAGCAA3’) with annealing temperature of 61°C was identified to be polymorphic between the parents. 10 randomly selected samples were taken from the resistant and susceptible plants to prepare bulks for bulk segregant analysis (Figure 1). The marker was found putatively linked to theLr28 locus. This polymorphic SSR marker was analysed on the 254 F₂ plants for linkage analysis with theLr28 locus. The marker Xwmc497 was associated with theLr28 locus and was located at a distance of 2.9 cM from it. The PBW343-Lr28 resistance allele linked SSR marker allele amplified a 291 bp fragment and the CSP44 susceptibility allele linked marker allele amplified a 226 bp fragment.

The 291 bp fragment was specific to theLr28 resistance allele and did not amplify in otherLr genes carrying lines from other native and alien sources.

By employing the flanking RAPD markers S3₄₅₀ (5’CATCCCCTG3’) and S336₇₇₅ (5’TCCCCATCAC3’) linked respectively to the recessive resistance allele and dominant susceptible allele of theLr48 locus; plants which were homozygous for recessive APR geneLr48 were identified, as these two markers served as one co-dominant marker system capable of identifying both dominant and recessive alleles of heterozygous plants. 70 F₂ plants were found to possess the homozygous recessive resistance allele ofLr48 out of the 254 plants (Table 1). Of these, only 14 plants were homozygous for the geneLr28 also and were identified to be carried forward as breeding lines.

Table 1: Mean grain yield of the F₂ plants pooled with reference to the segregation of the resistant alleles of the marker loci. ^§Codominant microsatellite marker; R: Homozygous resistant; ^¶Dominant RAPD marker; H: Heterozygous resistant; S: Homozygous susceptible; +: Presence of RAPD marker fragment; -: Absence of RAPD marker fragment.

The grain yield of each plant was recorded in order to advance only those which were comparable to PBW343 in mean yield/plant and displayed rust resistance imparted by bothLr28 andLr48 (Table 1). PBW343 is a high yielding Indian cultivar and had a mean single plant yield of 9.50 gm while the APR parent CSP44 recorded a lower yield of 8.78 gm. The mean yield of the 14 plants homozygous forLr28+Lr48 was 9.49 gm. These would be advanced as pyramided lines and followed for ear-to-row progeny analysis without elimination to select for high yielding recombinants through pedigree selection approach as the two genes are fixed in these progenies.

Gene pyramiding holds its base on the concept that the probability of mutation at more than one avirulence gene locus in the pathogen is low for it to turn virulent for all the pyramided resistance genes. This enables a host variety which possesses more than one gene to remain durably resistant to the disease relatively for a long period compared to the single gene based resistance. In addition, when the added gene is from wild species the resistance is expected to last long as matching virulence is less likely to be present in the pathogen population. Further, if the resistance is race non-specific such as APR, there would be still less chance for virulence development for all the prevailing races. Thus a pyramided combination of alien seedling resistance and APR would be an ideal means to ensure durable resistance. In the past three decades, combinations of alien and APR genes such asLr16 andLr13,Lr13 andLr34,Lr13 andLr37,Lr34 andLr37 have been achieved through conventional means as there were available pathogen virulence differentials or phenotypic differences in reaction types to distinguish each gene [19,20]. However, in a case where the presence of both genes cannot be detected due to lack of such differences as in the case ofLr28,Lr24, etc, a selection process which employs molecular markers tagged to the genes is a reliable methodology as has been demonstrated by in pyramidingLr24 andLr48 in wheat by marker assisted selection utilizing dominant SCAR and RAPD markers in consecutive generations till homozygosity was achieved at both loci [10]. We were able to identify plants fixed for both genesLr28 andLr48 in F₂ generation itself owing to the codominant SSR marker in combination with the flanking RAPD marker set linked to both recessive resistance and dominant susceptibility alleles at theLr48 locus. Gene pyramiding is well utilized in rice breeding programmes also to develop plants carrying Xa21 and xa13 resistant to bacterial blight which has also led to commercial release of the pyramided variety in India. Marker assisted pyramiding is also reported against fungal blast (Pi1 and Pi2) and brown plant hopper (Qbph1 and Qbph2) [21]. This strategy is being followed in many other breeding programmes with various crops for a range of beneficial phenotypes.

Seedling resistance genes such asLr28 are important to control the pathogen infection during the entire crop duration. There are previous reports of identified markers tagged toLr28. The SCAR marker SCS421₅₇₀ is being successfully employed in various wheat breeding programmes in India. A recent publication by has suggested the utility of two SSR markers, Xbarc327 and Xbarc343 to identify the presence ofLr28 [5,22]. However, these two markers were found to be monomorphic amplifying the critical marker fragment in both the parents. A null allelic microsatellite marker, Xgwm160 has also been reported to be specific to theLr28 gene. Xgwm160₁₉₆ and Xwmc497₂₉₁ are positioned at a distance of 144.9 cM and 149.9 cM respectively, from the centromere on the long arm of the 4A chromosome [6,14].

The microsatellite marker reported in this paper will be helpful for breeding purposes since it differentiates the presence of the gene in homozygous resistant and heterozygous resistant plants (Figure 2). It has been suggested by that the markers should be within 10 cM of the gene of interest for effective marker-assisted selection breeding [23,24]. The marker Xwmc497 mapped at a distance of 2.9 cM will therefore be especially useful for those breeding programmes in wheat where pyramiding is performed to stack more than one resistant gene into a single background. In the current study, molecular markers were effectively used to identify pyramided single plants in the F₂ generation itself which otherwise would have needed a laborious and time consuming selection process consisting a combination of phenotype based selection and a dominant marker based selection till the F₅/F₆ generations.

Figure 2: Segregation of the marker Xwmc497₂₉₁ in the heterozygous F₂ population. Individual F₂ plants amplifying the specific bands: Lanes 1, 3, 5, 6, 10, 13, 16, 18, 20, 21, 22, 24: heterozygous resistance, Lanes 2, 4, 8, 9, 11, 14, 17, 23: homozygous susceptibility, Lanes 7, 12, 15, 19: homozygous resistance; M: 100-bp DNA ladder.

The RAPD marker pair S3₄₅₀ and S336₇₇₅ which we used in the study had an advantage enabling us to successfully identify the plants which carried only the recessive adult plant resistance allele pair of theLr48 locus. From among 254 F₂ plants, we could select 14 plants carrying both the genes.

The grain yield of a plant follows a quantitative inheritance pattern and the expression of resistance is a qualitative character and there is no available information suggesting the influence of the leaf rust resistance loci on the grain yield of the plant. In this experiment we have also scrutinized the plants on the basis of their yield and only those plants with adequate grain number and with the presence of both the resistant genes were chosen. The 14 plants were comparable with PBW343 for mean yield/plant. The progeny of these plants will be carried forward through marker assisted pedigree breeding procedure.

The authors are grateful to the Indian Council of Agricultural Research for sponsoring the project and funding the fellowship to JK and AS under Molecular Breeding Network Project. We acknowledge Dr R. G. Saini for supplying the parental material of theLr gene donors. The authors are grateful to Head, Regional Station, Indian Agricultural Research Institute, Wellington for providing pure seed of the near-isogenic lines of wheat and Directorate of Wheat Research, Flower dale, Shimla for providing pure inoculums of leaf rust pathogen.