Multi-Pathotype Testing of Selected Kenyan Wheat Germplasm and Watkin Landraces for Resistance to Wheat Stripe Rust (Puccinia striiformis f. sp tritici) Races

Wheat stripe rust, caused by Puccinia striiformis f. sp. tritici (Pst), is one of the key diseases of economic importance in wheat worldwide. Host resistance, which follows the gene-for-gene hypothesis between the host and pathogen, has been used in wheat lines to resolve resistance specificities and postulate resistant genes. The objective of this study was to elucidate stripe rust resistance in a collection of Kenyan wheat lines and Watkin landraces to identify new sources of stripe rust (Yr) resistance. In this study, the resistance in twenty wheat lines was determined by comparing their infection type with those of twenty differential lines using isolates representing twelve Puccinia striiformis races from Kenya, Denmark, U.K., Sweden, and Eritrea at the seedling stage. Among the twenty wheat lines, none was resistant to all the twelve Pst races and isolate DK02d/12 (“Kranich” race) was virulent on all the genotypes except wheat genotype “Kenya Tai.” This genotype (“Kenya Tai”) had the highest resistance as it was resistant to all the twelve stripe rust races used in this study. From this study, the introduction and utilization of wheat genotypes with adult plant resistant (APR) stripe rust genes, such as Yr15, are important in breeding wheat genotypes with effective resistance to wheat stripe rust in Kenya and worldwide.


Introduction
Wheat (Triticum aestivum L.) is the second-most important cereal crop in Kenya, after maize; however, there is inadequate annual production due to abiotic and biotic factors leading to a deficit of over 60% [1]. Wheat stripe rust caused by Puccinia striiformis f. sp. tritici (Pst) is one of the biotic factors that has been an important disease in Kenya for the past century [2,3] and its pathogenicity increases yearly [2,4,5]. The epidemiology of stripe rust is influenced not only by the deployment of resistance genes in wheat varieties, but also by climatic conditions that affect Pst infection and growth and by wind movements across areas where wheat is grown. Among the countries in East Africa, stripe rust is important in Kenya, Ethiopia, and Uganda [6]; hence, regular virulence analysis is vital for disease management and the incorporation of resistant genes [7]. A stripe rust race (pathotype) is the compatible or incompatible interactions between the host and pathogen by characterizing the Pst on a set of resistant (R) genes carrying wheat differentials, and associated race surveys are undertaken in Table 1. Wheat (Triticum aestivum L.) genotypes and differential lines used for stripe rust (Puccinia striiformis (Pst)) virulence analysis.

Variety/Landrace
Year of Release Pedigree/Accession Name Seed Source

Stripe Rust Isolates
Stripe rust isolates used in this study were provided by Prof. M.S. Hovmøller of GRRC, Denmark. A total of twelve stripe rust isolates of diverse origins from five countries (Kenya, Denmark, U.K., Sweden, and Eritrea) were used ( Table 2). The isolates were selected due to their broad spectrum in terms of virulence, and because they represent different world regions that have been collected from for the last 40 years. Urediniospores used in this study had been previously (in a different study on virulence analysis, unpublished data) purified and developed on the susceptible wheat genotype "Cartago," and subsequently stored over liquid nitrogen (−196 • C) at GRRC's rust isolate bank in accordance with References [8,21,22].

Inoculation and Incubation of the Wheat Genotypes
Virulence assessment of the races on the resistance of the wheat genotypes based on their infection type (IT) was carried out on the Kenyan and Watkins lines, and differential sets. Seven seeds of each genotypes were sown in 6.5 cm × 6.5 cm square plastic containers filled with approximately 144 g of peat moss (PINDSTRUP MOSEBRUG A/S 8550 RYOMGARD, Denmark) and placed on a tray measuring 35 cm × 44 cm. The sown seeds were placed in a spore-proof greenhouse maintained at a temperature of 17 • C/12 • C day/night cycle. The stripe rust isolates were removed from the liquid nitrogen (−196 • C) and heat-shocked in a water bath at 40 • C for 2 min. At the two-leaf stage (feekes stage 1) [23], the wheat genotypes were inoculated with urediniospores (15 mg) of each isolate suspended in Novec TM 7100 using an air brush spray gun [24], then incubated in the dew chamber at 100% humidity and 10 • C for 24 h in total darkness. The inoculated plants were transferred to spore-proof greenhouse cabins with a 16-h/8-h day/night cycle, maintained at 17 • C/12 • C day/night cycle. susceptible to three types of rust, was planted perpendicular to all entries. The disease occurred naturally in all three environments.

Data Collection and Analyses
Infection types (IT) were noted 16 days after inoculation based on a 0-9 scale [25]. ITs of 0-4 were classified as resistant, ITs of 5-6 as intermediate, and ITs of 7-9 as susceptible. The resistance spectrum was based on a gene-for-gene hypothesis where the ITs produced by the Pst races on the wheat genotypes (Kenyan and Watkins lines) was compared with the ITs produced by the same races on the tester lines with known genes (differential lines) [16,26]. When the pathogen was avirulent, the response IT was classified as resistant; however, when the races were virulent on the host, the response IT was classified as susceptible and the intermediate reactions were classified as moderately resistant and moderately susceptible. The spectrum was comprised of resistant (R), moderately resistant (MR)/susceptible (MS), and susceptible (S) reactions. The disease in the field experiment was assessed during the peak of the wheat rust infection, immediately after heading. The host reaction and rust severity were taken from each entry. Estimates of disease severity were based on a modified Cobb scale [27,28] incorporating both the percentage of leaf area affected and the host response.

Confirmation of Virulence Phenotype of the Twelve Yr Isolates and the Response of the Differential Lines to the Isolates
Collectively, the twelve Pst isolates were virulent on the following yellow rust (Yr) genes: Yr1, Yr2, Yr3, Yr4, Yr6, Yr7, Yr8, Yr9, Yr10, Yr17, Yr25, Yr27, Yr32, YrAmb, and YrAvS, and avirulent to Yr15 and YrSp (Table 2). Isolates from Kenya and Denmark made up the majority of virulent isolates, consisting of 41.67% and 33.33%, respectively. The Kenyan isolates had a common virulence for yellow rust genes Yr2, Yr6, Yr7, and YrAvS, except KE86102Yr9, which lacked virulence for Yr7 and Yr8, but was the only Kenyan isolate with virulence for Yr17. Virulence for Yr17 was present in the European isolates. Two pathotypes, KE70063 and KE131/14, had additional virulence for Yr25 and Yr27; moreover, the latter also had virulence for Yr1 and Yr9. The isolate from Eritrea had virulence for Yr2, Yr6, Yr7, Yr8, Yr9, Yr25, YrAvS, and an additional virulence for Yr10 that was absent in the Kenyan isolates ( Table 2). The Danish isolates (DK01/09; DK71/93) had virulence for 6 Yr genes, whereas isolates (DK09/11SP; DK02d/12) had virulence for 12 Yr genes, thereby representing isolates with a wide virulence spectrum. The other isolates from Europe (Sweden and United Kingdom) differed in virulence, SE100/09 being virulent on only five Yr genes and UK94/519 had virulence for seven Yr genes. Interestingly, the Swedish isolate and one Kenyan isolate were avirulent to YrAvS.
a and b are infection type (ITs) and the reaction of genotype to the stripe rust isolates. R (resistant); MR (moderately resistant); MS (moderately susceptible); S (susceptible). ITs are given according to the scale of 0-9 [25]. The resistance response spectrum was according to Hovmøller [16].   a and b are infection type (ITs) and the reaction of genotype to the stripe rust isolates. R (resistant); MR (moderately resistant); MS (moderately susceptible); S (susceptible). ITs are given according to a scale of 0-9 [25]; The resistance response spectrum was according to Hovmøller [16].

Genotype
Pedigree Gene postulation was based on the resistance spectrum from the seedling results in Tables 3 and 4. Resistance genes of the genotypes were postulated by comparing resistance spectra of the genotype to twelve races with those of wheat genotypes possessing known Yr resistance genes (differential lines) [9]. R (resistant); MR (moderately resistant); MS (moderately susceptible); M (moderately resistant to moderately susceptible); S (susceptible). Disease severity was based on a modified Cobb scale [28].

Confirmation of the Virulence Phenotype of the Twelve Yr Isolates and Response of the Differential Lines to the Isolates
The Pst isolates used had virulence for Yr1, Yr2, Yr3, Yr4, Yr6, Yr7, Yr8, Yr9, Yr10, Yr17, Yr25, Yr27, Yr32, YrAmb, and YrAvS, but were avirulent to Yr15 and YrSp. There was virulence for Yr1, Yr2, Yr6, Yr7, Yr8, Yr9, Yr17, and YrAvS on Pst isolated collected before the year 2000 and after the year 2000 in Kenya (unpublished data). Virulence for Yr17 has not been common in Kenya; however, most of the European pathotypes were virulent on Yr17; this could be due to the presence of Pst1 and Pst2 strains, which are also present in East Africa [29]. The varieties deployed in these countries, such as "Bill," "Biscay," and "Bandit" from Denmark and "Equinox" from England have been postulated to have Yr17 [16].
The difference in virulence for pre-2000 and post-2000 Danish isolates is due to the evolution of the new races, such as the "warrior" race, that was detected in Europe in 2011 infecting many wheat varieties and triticale (Tritico secale), where the isolate DK09/11SP used in this study belongs to the warrior race, which is virulent to Yr2, Yr6, Yr7, Yr8, and Yr10 [21]. From the results, we can group the isolates' pathotypes into two different groups according to the regions of East Africa (Eritrea and Kenya) and Europe (Denmark, Sweden, and U.K.), which brings out the geographical distribution of Pst, and the Middle East and East African population were grouped together as Middle East-East African as they did not differ much from each other [29]. Pst1 and Pst2 were present in East Africa (Kenya) and the same strains were present in Europe in addition to the "warrior" race. This brings out the evolution and similarity in Pst strains, which has been observed in Europe (Denmark) and East Africa (Kenya and Ethiopia), both from different continents with common virulences for Yr2, Yr6, Yr7, and Yr8 [21,30,31]. Wheat genotypes with resistant genes Yr2, Yr6, and Yr7 have been extensively used in Kenya and the same virulence to these genes were also found in isolates obtained from Hungary and Canada [29]. The pathotypes from different countries shared common virulence but differed in the degree of their virulence due to differences in the aggressiveness of the races against the deployed resistant genes in the regions [32].
Gene pyramiding is effective in conditioning resistance as "CarstensV" with resistant genes (Yr32, Yr25) was resistant to more isolates than "TP981" (Yr25). The difference in resistance of Yr17 in "VpM1" and in "AvocetS" was the same as the resistance to Yr6 in the "Heines Kolben" and "AvocetS" background, which brings out the interaction of genes in different background conditioning differences in gene expression and resistance.
Virulence to Yr5 and Yr15 were not detected in this study. This is because Yr5 and Yr15 are resistant to a wide range of isolates of yellow rust from different geographical regions [26]. Therefore, these genes, in combination with other race non-specific genes, can be deployed in wheat breeding programs to condition resistance [27,30].
Based on the multi-pathotype tests conducted, the wheat lines could have had different stripe rust resistance genes, with some similar to the tester lines and others being unknown. Yellow rust genes Yr2, Yr7, Yr9, and Yr15 were postulated in the Kenyan line tested. "K. Korongo" and four Watkins lines ("1190713," "1190026," "1190826," "11900034," and "1190524") did not match any spectra of the testers, indicating the presence of unknown resistance genes in these lines.
Stripe rust Yr1 gene was not postulated in the wheat lines tested in this study; however, the same gene was postulated in Chinese landraces "Zhengmai366," "Xinmai208," and "Luomai21", and in cultivar "Buster" [9]. Gene Yr2 was postulated in several studies in Nepal, Europe, and China [17,27,33]. The stripe rust gene (Yr2) is no longer effective against the Yr pathotypes tested and it might be present in wheat varieties "1190120," "K. Wren," "K. Hawk12," "K. Sunbird," and "Robin," which explains why these varieties were susceptible to most of the pathotypes used. This gene (Yr2) found in the differential line "Kalyansona" is no longer effective; therefore, the lines with a single Yr2 need to be improved by incorporating resistant genes such as Yr15, YrSP, Yr4, Yr10, and YrAmb to attain durable resistant.
Gene Yr6 was postulated in Australia wheat lines categorized under group six (YRG6) [28] and in wheat cultivars "Hunter," "Lynx," and "Rialto" [9]. In our study, no genotype was postulated to have Yr6. Gene Yr7 could be present in "Duma" and 'K. Kingbird," where this gene is linked to Sr9g in cultivar "Cadenza" and "Tonic" [34]. Gene Yr9 postulated in "Kwale" and "NJBW2" is pleotropic or linked to Sr31/Lr26. However, virulence to Sr31 has been present in Kenya since the emergence of Sr31 virulence in Uganda [9]; therefore, the gene is ineffective and cannot be deployed in Kenyan wheat lines unless in combination with other resistant genes. Genotype "K. Tai" was postulated to have Yr15 plus other unknown genes. Yr15 is a T.-dicoccoides-derived stripe rust resistant gene with all stage resistance (ASR) located on chromosome 1BS [35] and was found in 25 wheat lines from Iran [20]. The genotypes that had an unknown gene postulated include all the Watkins lines except "1190120," exhibiting more than 50S in either one of the locations tested, which is an indication of the lack of effective genes against the pathotypes present in that environment.
The effectiveness of stripe rust genes against the new Pst pathotypes is not clear and the genetic background plays a big role in this; hence, it is important to have a diverse germplasm for testing [36][37][38]. Yr15 and Yr65 are among the genes that are still effective, while Yr9 and Yr10 are among the ineffective genes [39].

Conclusions
Evaluation of Kenyan and Watkins lines with twelve races from Kenya, Eritrea, Denmark, Sweden, and U.K. indicated that race DK02d/12 from Denmark was more virulent on the genotypes than any other race except genotype "K. Tai," which was resistant to these races and others; hence it is useful for breeding wheat cultivars with resistance to stripe rust. The twenty wheat genotypes that exhibited similar infection type responses as the tester lines could have had the same genes as the tester lines, but this needs further research to confirm the genes present.