Introduction
In various developing countries, wheat is an essential crop equally in consumption and production. On more than 200 million hectares of area, the wheat is cultivated and the total production of wheat is about 733 million tons every year (FAO, 2015). Since the world population is increasing fast and predictable to touch over 9 billion in 2050 (Edmeades et al., Reference Edmeades, Fischer and Byerlee2010), the production of more wheat is likely to become a great urgency (Wise, Reference Wise2013).
Wheat production is influenced by the stripe, stem and leaf rusts. Among these, stripe rust (caused by Puccinia striiformis Westend f. sp. tritici Erikss) is specified by high frequency, a wide area of incidence and large economic damages is one of the most predominant wheat diseases, responsible for great production losses globally (Chen et al., Reference Chen, Wellings, Chen, Kang and Liu2014). Presently, about 88% of the worldwide wheat production is prone to wheat stripe rust, responsible for worldwide damages of more than 5 million tons of wheat (Beddow et al., Reference Beddow, Pardey, Chai, Hurley, Kriticos, Braun, Park, Cuddy and Yonow2015).
Stripe rust resistance is generally characterized as all-stage resistance (ASR) and adult-plant resistance (APR) (Chen, Reference Chen2005). Among these two, ASR is shown in all plant growth steps and might be identified in seedlings. ASR resistance is usually ideal in breeding for the reason that the specific genes give high levels of resistance. This type of resistance is exposed more to the evolution of pathogen when used individually in cultivars (Chen et al., Reference Chen, Penman, Wan and Cheng2010). On the other hand, APR normally gives defence against a wider range of races and inclines to be durable (Niks et al., Reference Niks, Qi and Marcel2015). Until now in wheat, more than 80 yellow rust (also known as stripe rust) resistance (Yr) genes have been designated, together with the newly mapped Yr79 (Feng et al., Reference Feng, Wang, See, Chao, Zheng and Chen2018) and Yr80 (Nsabiyera et al., Reference Nsabiyera, Bariana, Qureshi, Wong, Hayden and Bansal2018). Though these Yr genes have been found in various wheat accessions, the seedling resistance gene's race specificity confines their effectiveness against pathotypes. In contrast, APR is commonly regarded to be durable, but APR genes denote a smaller number of known resistance genes (Kankwatsa et al., Reference Kankwatsa, Singh, Thomson, Babiker, Bonman, Newcomb and Park2017). Bux et al. (Reference Bux, Ashraf and Chen2012) found Pakistan landraces with high-temperature adult plant resistance (HTAP) resistance. The accessions having HTAP resistance worked as new germplasm for breeding resistant cultivars for wheat-growing areas of Pakistan such as southern Sindh and Punjab, where high-temperature climates are appropriate for the presence of such kind of resistance against stripe rust.
The use of cultivars using single-gene resistance (race-specific resistance) allows selection of mutations on a single locus. Because of selection pressure, novel virulent races of fungus arise, therefore, the utilization of mixtures of resistance genes has been proposed as the finest way for genetic control of stripe as well as other rusts. Slow rusting provides more durable resistance against stripe rust (Venkata et al., Reference Venkata, Bansal, Singh, Park and Bariana2008). The current study was carried out to characterize the recombinant inbred line (RIL) population developed from a landrace originated from Pakistan for rust resistance. The slow rusting or APR mechanism in the landrace was characterized at the phenotypic level to demonstrate the usefulness and inheritance of putatively new rust resistance gene.
Materials and methods
Plant material
A total of 25 landraces was selected randomly on the basis of field reactions 2014–2015, the details of 25 landraces showing the collected sites from different areas of Pakistan is shown in Table 1. Under glasshouse conditions, a set of 25 landraces was screened against stripe rust races 574730, 574212, 574232, 574216, 410202, 430220 and 476232 at seedling stage (Table 2). A mapping population was developed by a cross between landrace PI388222 and susceptible Australian spring wheat line, Avocet ‘S’. Landrace PI388222 showed moderately resistant to resistant reaction to stripe rust in field conditions, so it was taken as a resistant parent (female) and Avocet ‘S’ was selected as a susceptible parent (male). The seed of both parents was kindly provided by USDA small grain collection Aberdeen, Idaho, USA and conserved at CDRI Murree seed collection. A total of 136 lines were used for greenhouse and field stripe rust phenotyping.
List of 25 landraces collected from different areas of Pakistan and evaluated for stripe rust resistance in this study. PI refers to plant identification number at National Small Grain Genebank, Aberdeen, Idaho, USA
Disease score of 25 wheat landraces against stripe rust resistance at seedling stage against seven P. striiformis races
The F1 produced were sown separately in small pots under glasshouse conditions at Crop Disease Research Institute (CDRI) Murree. The F2:4 families were developed in the field of National Agricultural Research Centre (NARC), Pakistan, from single seed descent from F1plants. The F2:4 families were sown and assessed for stripe rust resistance to natural inoculum in 2016 at NARC (Pakistan). From the F2:4 seeds collected from the field, single seeds were planted in the greenhouse of Washington State University Pullman, WA to develop F4:5 RIL. The F4:5 lines were used for additional phenotyping.
Phenotyping in the greenhouse
Seedling test
The F2 and F3 of the population (PI388222 × Avocet ‘S’) were screened for stripe rust resistance at the seedling stage. The seedling screening was conducted in greenhouse conditions in CDRI Murree. Murree is a hilly resort town, situated in the Galyat region of the Pir Panjal Range, in the District Rawalpindi of Punjab, Pakistan. It makes the borders of the Islamabad-Rawalpindi metropolitan zone and is nearly 30 km (19 mi) northeast of Islamabad. Nearly five seeds of each line were sown in pots (7 × 7 cm2). The inoculums which are conserved at −80°C temperature were taken from the freezer and rapidly dipped in water bath, the temperature of which was fixed at 65°C for 15 min for breaking the dormancy of spores. The inoculums suspension was made in the mixture of petroleum ether and paraffin oil (80:20 v/v) and sprayed on 14 d old plants with the help of fine atomizer at two leaf period (Rizwan et al., Reference Rizwan, Iftikhar, Kazi, Sahi, Mirza, Attiq-ur-Rehman and Ashraf2010). Prior to shifting the plants to growth room whose temperature was fixed at 10°C, the plants were exposed in air for 2 hours and at comparative moisture of 100. A 12-hour period was provided to plants in a growth room, after this period they were moved to glasshouse at 18–20°C temperature. Three weeks after inoculation, infection types (ITs) were noted by the help of 0–9 scale (Line and Qayoum, Reference Line and Qayoum1992) when higher infection was exhibited by Morocco (susceptible check). Plants displaying infection type of 0–4 were measured as resistant, whereas plants revealing ITs 5–6 were categorized as intermediate resistance, and plants displaying infection type of 7–9 were measured as susceptible. The seedling test was done on F2 stage for the race (574212) and at F3 the races used were 574212, 476232, 430220 and 574232. The seedling test done on F5 stage was against races (PSTv-37, PSTv-40, PSTv-4, PSTv-51) in the glasshouse of Washington State University, Pullman USA. Seedling screening process in WSU, Pullman USA was dissimilar. It comprised the following steps: On the leaves of Avocet ‘S’ (susceptible spring wheat line), fresh urediniospores were produced as specified by Chen et al. (Reference Chen, Line and Jones1995). After about 16 d of inoculation, the collection of fresh urediniospores was done three to five times for enough quantity. Collected urediniospores were desiccated in a desiccator at 4°C and utilized for inoculation for 1-month period. Sowing of RIL (F5:6) and parental seeds was done in 72-well trays having five seeds/ line and developed in the greenhouse. Separate inoculation was done for seedlings at two-leaf stage (14 d after planting) and adult-plants at booting stage with urediniospores of every single race mixed with talc nearly 1:20 ratio (Line and Qayoum, Reference Line and Qayoum1992). The seedling incubation was done in a dark dew chamber, inoculated seedlings were incubated at a temperature of 10°C for around 24 h and shifted to a growth chamber having a diurnal temperature cycle gradually varying between 4°C at 2:00 am and 20°C at 2:00 pm and 8 h dark/16 h light as designated by Chen and Line. On inoculated seedlings, ITs were noted at 18–21 d after inoculation depending on the 0–9 IT scale for stripe rust.
Field screening for populations (PI 388222 × Avocet 'S') on F5 stage
Field screening of the population was conducted in USA at two locations i.e. Palouse Conservation Field Station (PCFS) field and Mount Vernon (MV) during the wheat growing season of the year 2017. Inoculations of border rows at Pullman were done on 15 June 2017 with the race PSTv-37. Border rows were planted with Avocet ‘Susceptible’ (Avocet ‘S’). All spring wheat plots were planted in PCFS on 10 May 2017. The following day (11 May), 100-pounds/acre fertilizer (urea, 46-0-0) was applied. In MV, the populations were exposed to natural infection. For stripe rust resistance, RILs and parental lines were screened in the year 2017. Seed sowing was carried out in rows of 0.5 m set apart at 30 cm. In 2017, three replications were sown in randomized complete block design (RCBD) at both locations. The Avocet ‘S’ was planted around the plots as spreader rows and was planted after each 20th row to improve even rust development all over the field. For the F5 RIL population, DS was changed to AUDPC values according to Chen et al. (Reference Chen, Line and Jones1995). For the parents and every RIL of population, AUDPC was measured 
$^{\prime}{\rm AUDPC} = \sum _i[(x_i + x_i + 1)/2]t_i^{\prime}$. For each RIL and parent, relative AUDPC (rAUDPC) values were calculated as a % of mean AUDPC value of the susceptible parent, Avocet ‘S’ (Lin and Chen, Reference Lin and Chen2007).
Statistical analysis
For the assessment of the goodness of fit of ‘observed to expected segregation ratios’, χ 2 tests were used to assess the number of resistance genes adapting resistance.
Results
Glasshouse results of PI388222 × Avocet ‘S’ at F5 stage
The response of the 25 landraces against Pakistani PSTv races is shown in Fig. 1 and the avirulence/virulence formula of the races is shown in Table 3. Landrace PI388222 gave infection type 4 with race 574212, infection type 5 with 574232, infection type 3 with race 430220 and infection type 4 with 476232. The response of Landrace PI388222 and near-isogenic lines to selected Pst races is shown in Table 4. Among the 7 PSTv races, the maximum resistant response shown by the landraces was against the race 430220 (18R and 7MR and none of the landrace exhibited MS-S reaction), whereas the minimum resistant response of the landraces was shown against the PSTv race 574216 (3R, 14MR and 8MS) as shown in online Supplementary Table S1. A variable set of seedling reactions were detected when a set of 136 RILs together with parental lines were screened with Puccinia striiformis f. sp. tritici races PSTv-37, PSTv-40, PSTv-4 and PSTv-51 for identification of the putative major gene. The virulence/avirulence formulae for PSTv races is shown in online Supplementary Table S2. The seedling response of RIL population to the PSTv-37 identified 67 (49%) RILs were resistant, 57 (42%) were moderately resistant and 12 RILs (9%) were susceptible and the response of parents was 3 and 8 for PI38822 and Avocet ‘S’, respectively (Fig. 2). The seedling response of RIL population to the PSTv-40 revealed that out of 136 RILs, 116 (85%) were resistant, 19 (14%) were moderately resistant, and only one (1%) RIL was susceptible (online Supplementary Fig. S1). The PI38822 response was 3 and scored as moderately resistant, while Avocet ‘S’ response 8 and scored as susceptible. The score of all RILs against PSTv-40 was R/MR or less than 5 because the selections were made on the basis of their reaction to Pst race 574212 and were raised to F5. Then the material was tested in Pullman, USA, where it was screened against different stripe rust race. The selection process would have skipped lines susceptible to US stripe rust race (PSTv-40), as the race was not available during the selection process. The seedling response of RILs population to the PSTv-4 showed that 102 (75%) RILs were resistant, 26 (19%) were moderately resistant and 8 (6%) were susceptible (PI388222 response was 3 and response shown by Avocet ‘S’ was 9) (online Supplementary Fig. S2). The seedling response of RIL population (PI38822 response was 3 and Avocet ‘S’ was 9) at F5 to the PSTv-51 revealed that out of 136 RILs 114 were resistant, 18 RILs were moderately resistant and 4 RILs were susceptible (online Supplementary Fig. S3). The maximum resistant response shown by the RILs was against PSTv-40 against which 85% of the RILs were resistant while the low resistance to the race PSTv-37 was noticed against which 49% of the RILs were resistant.
Reaction of landraces to Pakistani Pst races 574730, 574212, 574232, 574216, 410202, 430220 and 476232.
Frequency distribution of infection types in F5 population of the cross PI388222 × Avocet ‘S’ when screened with PSTv-37 at F5.
Avirulence/virulence formula for stripe rust
Response of Landrace PI388222 and near-isogenic lines to selected Pst races
LITa, Low Infection Type.
Adult plant resistance in PI388222 × Avocet ‘S’ RILs
Frequency distribution of mean IT and DS in PCFS and MV are shown in online Supplementary Figs S4–S7. Higher disease pressure was observed at both sites and maximum rust severity up to 100% was observed in susceptible check ‘Morocco’. The FRS of the RIL population and the parents (PI388222 and Avocet ‘S’) is described in online Supplementary Table S3. In RILs population, 80% of the RILs were resistant in PCFS, 12% showed moderate reaction and 8% showed susceptible reaction. The disease severity was higher in MV, where 76% of RILs were resistant, 15% were moderately resistant, 9% were susceptible. Out of 136 RILs, 53% of the RILs revealed resistant reaction towards four Pst races that are used in glasshouse screening and were also resistant in field environments at PCFS and MV.
The rAUDPC was used for identification of slow rusting or durable resistance germplasm. At PCFS, 42 (31%) RILs were resistant with rAUDPC 0-10, while 70 (51%) RILs had rAUDPC values of >30 and were intermediately resistant, whereas 24 (18%) RILs had rAUDPC of 30-90 and were susceptible. The MV field results showed that only 1 RIL was resistant, 12 RILs (11.5%) were intermediately resistant, and 91 RILs were (87.5%) were susceptible based on rAUDPC (online Supplementary Table S3).
Discussion
The genetic resistance is a cost-effective technique for the development of resistant varieties. One of the best strategies for finding new resistant sources and the discovery of durable rust-resistant genes is the characterization of the primary gene pool of wheat (Mujeeb-Kazi et al., Reference Mujeeb-Kazi, Kazi, Dundas, Rasheed, Ogbonnaya, Kishii, Bonnet, Wang, Xu, Chen, Mahmood, Bux and Farrakh2013). Primary gene pool includes landraces, wild and primarily domesticated wheat relatives. Landraces are normally considered to have good diversity for biotic and abiotic stress tolerances (Zeven, Reference Zeven1998). Wheat landraces are favourable source of new genes of rust resistance for developing novel and diverse resistant germplasm (Sthapit et al., Reference Sthapit, Newcomb, Bonman, Chen and See2014).
The set of 25 wheat landraces showed a high resistant reaction in the field which indicated that these landraces may have some important resistant genes. The response of a selected wheat landrace PI38822 against stripe rust was moderately resistant to resistant in the field, which indicated the presence of slow rusting genes, and was selected for further study.
Previously, a wheat landrace collected in Pakistan, AUS27858, from the Watkins collection showed high levels of resistance against Australian pathotypes of Pst. An analysis was carried out on AUS27858/Westonia F3 population showed that AUS27858 possessed two stripe rust resistance genes that were temporarily designated as YrAW1 and YrAW2 (Randhawa et al., Reference Randhawa, Bansal, Valárik, Klocová, Doležel and Bariana2014). Randhawa et al. (Reference Randhawa, Bansal, Valárik, Klocová, Doležel and Bariana2014) further developed an F6 RIL population and identified that YrAW1 is located on the long arm of chromosome 4A, and officially designated as Yr51 (https://maswheat.ucdavis.edu/protocols/Yr51/index.htm). Segregation of resistance identified in segregating population of PI388222 × Avocet ‘S’ in F2s and F3s revealed a good fit to the expected segregation ratio (15:1) for two genes, except PSTv-51 which indicated the presence of an extra gene against this race. Previously, Chhuneja et al. (Reference Chhuneja, Kaur, Garg, Ghai, Kaur, Prashar, Bains, Goel, Keller, Dhaliwal and Singh2008) carried out an experiment on RIL population produced from the cross of T. monococcum (acc, PAU 14087) and T. boeoticum (acc, pau5088) and identified the presence of two genes for stripe rust resistance.
A variable set of seedling reactions were observed when a set of 136 F2:F5 RILs of the cross PI388222 × Avocet ‘S’ were screened with races PSTv -37, PSTv -40, PSTv -4 and PSTv -51 for the presence of major genes. The Yr6, Yr9 and Yr27 were virulence genes against all the races mentioned above, excluding the chance of the existence in the landrace (Wan et al., Reference Wan, Chen and Yuen2016). Depending on avirulence pattern of pathotype, it is likely that one or combination of the stripe rust-resistant genes Yr10, Yr15, Yr24/Yr26, Yr44 could be present in PI388222. The Yr10 gene which normally gives low infection type with most of the races was identified in T. spelta (Bariana et al., Reference Bariana, Brown, Ahmed, Khatkar, Conner, Wellings, Haley, Sharpe and Laroche2002). Therefore, its presence in PI388222 could be ruled out. Similarly, Yr15 was identified in T. dicoccoides (Gerechter-Amital and Grama, Reference Gerechter-Amital and Grama1974), therefore it is likely to be absent in the landrace PI388222. Yr24 also known as Yr26 is identified in the synthetic hexaploid wheat contributed by the wild accessions of Ae. tauschii (McIntosh and Lagudah, Reference McIntosh and Lagudah2000). Therefore, it is likely that landraces, PI388222, does not contain Yr24/Yr26 gene. Yr44 is identified in cultivar Zak (released by Washington State University in 2000) (Cheng and Chen, Reference Cheng and Chen2010). Hence, the landrace PI388222 most likely does not contain Yr44, and genes present in this landrace could be new.
In earlier studies, field assessment of quantitative stripe rust resistance was used to classify RIL populations for adult plant resistance (Safavi, Reference Safavi2012). According to the results for field assessment, RILs showed the possibility of APR in the contender landrace together with the chance of major genes. Ali et al. (Reference Ali, Shah, Raman, Maqbool and Ullah2009) and Safavi (Reference Safavi2012) did field assessment for grouping of lines, which are used for quantitative resistance to stripe rust. Lines that displayed resistance reaction at both stages might more possibly have the major gene or mixture of major genes-based resistance effective for all utilized virulences. Because slow rusting resistance is durable, the study of the genetic nature of such type and its introduction to breeding material is striking for breeders (Zhang et al., Reference Zhang, Jun-Yi and Yong-Hua2003). On the basis of rAUDPC values, accessions were classified to two diverse groups: accessions showing slow rusting behaviours are likely to have genes that give partial resistance which is durable compared to the other resistance types (Parlevliet, Reference Parlevliet1988). The 31% of the RIL population at PCFS and 1% of RIL population at MV field showed lower rAUDPC and could have slow rusting genes.
Previously, cloning of the wheat multipathogen APR genes Lr34/Yr18/Sr57/Pm38 and Lr67/Yr46/Sr55/Pm46 shown that a single gene can give resistance to multiple pathogens (Moore et al., Reference Moore, Herrera-Foessel, Lan, Schnippenkoetter, Ayliffe, Huerta-Espino, Lilemmo, Viccars, Milne, Peiryannan, Spielmeyer, Talbot, Bariana, Patrick, Dodds, Singh and Lagudah2015). It is shown that these genes are involved in slow rusting at the adult plant stage and provide durable resistance against multiple fungal pathogens. Similarly, another gene Yr71 provides APR gene and can be combined with other major or minor stripe rust resistance genes to achieve low rust severity and durability (Bariana et al., Reference Bariana, Forrest, Qureshi, Miah, Hayden and Bansal2016). Durable resistance should be measured by investigators because the rust pathogens are capable of changing their genotypes without complications through migration, selection consequence of resistant cultivars on pathogens and mutation (Hovmøller, Reference Hovmøller2001). Several researchers similarly verified the presence of variable slow rusting levels in wheat breeding resources (Shah et al., Reference Shah, Hussain, Ahmad and Ibrahim2014). Formerly, Ali et al. (Reference Ali, Shah and Ibrahim2007) assessed wheat breeding material from Pakistan to evaluate their partial resistance regarding slow rusting behaviour. Many investigators employed FRS as a feature to measure slow rusting behaviour of wheat lines (Shah et al., Reference Shah, Hussain, Ahmad and Ibrahim2014).
Conclusion
The partial resistance identified in landrace PI388222 is due to the two genes or interaction between two genes (except for PSTv-51 which shows segregation of three genes). Based on the virulence pattern of Pst races, this landrace could have putatively new slow rusting genes. Further characterization of stripe rust resistance and mapping of resistance loci in this landrace will be valuable for developing new wheat cultivars for sustainable control of stripe rust.
Supplementary material
To view supplementary material for this article, please visit http://dx.doi.org/10.1017/S147926211900039X.
Acknowledgements
The authors are thankful to Dr Awais Rasheed for comments on the manuscript. The authors are grateful to Crop Disease Research Institute (CDRI) Murree and National Agricultural Research Centre (NARC) Islamabad, Pakistan for support of the studies embodied in the paper.
