Characterizing the pathotype structure of barley powdery mildew and effectiveness of resistance genes to this pathogen in Kazakhstan

Background Powdery mildew of barley is a wind-borne and obligate biotrophic pathogen, which ranks among the most widespread barley pathogens worldwide. However, purposeful research towards studying the structure of the barley powdery mildew populations, of their virulence and of effectiveness of certain resistance genes against the infection was not conducted in Kazakhstan till present time. This paper is the first to describe characteristics of the pathotype structure of Blumeria graminis f.sp. hordei (Bgh) population and effectiveness of resistance genes in two regions of barley cultivation in the republic. Results One hundred and seven isolates of Bgh were obtained from seven populations occurring on cultivated barley at two geographically locations in Kazakhstan during 2015 and 2016. Their virulence frequency was determined on 17 differential lines Pallas. All isolates were virulent on the resistance gene Mla8 and avirulent for the resistance genes Mla9, Mla1 + MlaAl2, Mla6 + Mla14, Mla13 + MlRu3, Mla7 + MlNo3, Mla10 + MlDu2, Mla13 + MlRu3 and Mlo-5. The frequencies of isolates overcoming the genes Mla3, Mla22, Mlat Mlg + MlCP and Mla12 + MlEm2 were 0.0–33.33%, and frequencies of isolates overcoming the genes Mlra, Mlk, MlLa and Mlh ranged from 10.0 to 78.6%. Based on reactions of differential lines possessing the genes Mla22, Mlra, Mlk, Mlat, MlLa and Mlh, pathotypes were identified. In total, 23 pathotypes with virulence complexity ranging from 1 to 6 were identified. During both years in all populations of South Kazakhstan and Zhambyl regions pathotypes 24 and 64 mainly prevailed. Conclusions Obtained data suggest that low similarity of populations Bgh in Kazakhstan to European, African, Australian and South-East Asian populations. The present study provides a foundation for future studies on the pathogenic variability within of Bgh populations in Kazakhstan and addresses the knowledge gap on the virulence structure of Bgh in Central Asia. Complete effectiveness of the resistance genes, for which no corresponding virulence was found, will allow Kazakhstanean breeders to access many modern barley cultivars that those possessing the resistance effectiveness genes.


Background
Powdery mildew caused by the biotrophic fungus Bgh (syn. Erysiphe graminis f.sp. hordei), is one of the most destructive diseases of barley (Hordeum vulgare L.) worldwide [1][2][3]. The disease is especially prevalent in moderate to temperate growing regions where yield losses can reach 40% [4]. Powdery mildew has a number of characteristics that support rapid evolution, such as large numbers of asexual haploid spores, sexual recombination during the growing season, and airborne dispersal over large distances [5].
Barley is the second after wheat most important for Kazakhstan cereal crop with total annual grain yield of more than 2 mln tons per hectare. Spring two-rowed barley represents over 90% of all barleys in the country [6]. Such a low output can be explained by strong pressure from abiotic factors, such as drought, heat, and heavy rains in autumn, and by periodic invasions of devastating barley pathogens [7]. Powdery mildew is one of the most powerful factors affecting barley production in Kazakhstan. Annually the pathogen affects crops of winter and spring barley in Kazakhstan. In recent years, disease epidemics have often been observed in the southern and south-eastern regions [8,9], indicating an expansive spread in the country. Most of the cultivated commercial varieties of barley lack sufficient resistance to powdery mildew and thus, the disease is of economic importance particularly in locations where conditions are conducive for disease development [9]. Breeding of resistant cultivars of barley is still impeded by high intrapopulation variability of Bgh and by the ability of the pathogen to overcome host resistance [1][2][3]. Consequently finding effective and durable control measures to constrain powdery mildew fungi represents an important challenge in crop protection research.
There are several ways of controlling the disease. The primary one is the use of genetically mildew-resistant cultivars. This is a cheap and environmentally safe method. There is a large number of mapped resistance genes that could provide protection against barley powdery mildew infection. More than 85 race-specific resistance genes for powdery mildew have been identified in barley [1]. Many near-isogenic lines have been developed by backcrossing different resistance gene donors into common genetic background, resulting in, e.g., the Pallas differential set [10]. Mla, one of the genetically most thoroughly characterized race-specific loci conferring resistance to powdery mildew, spans 32 known alleles and represents a true allelic series on chromosome 1H [11]. The Mlo resistance is a kind of resistance that is unique primarily because it is monogenic [12], although it does not conform to the gene-for-gene system [13], and also because its resistance mechanism differs substantially from that of the other kinds of resistance [1,12].
Powdery mildew can also be controlled with fungicides but these are ecologically undesirable and their frequent use may speed up the evolution towards resistance to fungicides [14]. The efficacy of using genetically mildewresistant cultivars depends on pathogen virulence [15]. Additionally, the continuous use of these genes often results in selection in favor of pathotypes with the matching virulence genes in the pathogen population, and therefore in the breakdown of the resistance [16].
Globally, populations of Bgh have changed and new and virulent pathotypes have emerged [17][18][19][20][21][22][23]. Individual genes of specific resistances differ substantially in their effectiveness against the pathogen population comprising both virulent and avirulent pathotypes [24]. New pathotypes not only cause severe epidemics but substantially limit production of commercial cultivars [17][18][19][20][21][22][23]. Despite the increasing importance of powdery mildew in Kazakhstan, the population structure of Bgh in this region has not been characterized and so far there were no attempts to determine effectiveness of powdery mildew resistance genes for future application in local breeding programs. This study was therefore undertaken to characterize the virulence structure of populations of Bgh in barley-growing regions and identify effective powdery mildew resistance genes in the Republic of Kazakhstan.

Pathotypes found
In years 2015 and 2016 23 pathotypes of the pathogen were detected among 107 isolates of Bgh with the help of the set including six differential cultivars (Table 3) (14) were original, 4 pathotypes were found in two populations, 1 pathotype in three populations, 2 pathotypes in four populations and 1 pathotype in five populations. Virulence complexity of pathotypes varied from 1 to 6. Wide  Calculated as number of isolates in each district as the proportion of the total number of isolates expressed as a percentage virulence range (virulence complexity equal to 5-6) was mainly characteristic to unique pathotypes (72, 57, 60) identified only in some populations and represented by single isolates (Tables 3 and 4). In South Kazakhstan and Zhambyl regions 9 pathotypes of the pathogen were prevalent (Fig. 3 (Fig. 3).

Descriptive parameters of populations
Parameters of intrapopulation diversity and statistical indices of difference between assayed populations of Bgh are shown in details in Table 5. In total 19 pathotypes were detected in populations of South Kazakhstan region, and nine in populations of Zhambyl region. The

Discussion
Prior to this study, pathogenic variability of Bgh populations has been investigated in a number of countries around the globe [17][18][19][20][21][22][23]. However, the virulence Based on 107 isolates collected during a 2 year period, 23 pathotypes with virulence complexity ranging from 1 to 6 were identified (Table 3). Two pathotypes, 24 and 64, were identified to be of importance in the barley production in Kazakhstan due to their occurrence across the entire study area. They differ only by one virulence gene, i.e. pathotype 24 is avirulent to gene Mlk, and pathotype 64 is virulent. In Kazakhstan all tested isolates were avirulent on lines with genes Mla9, Mla1 + MlaAl2, Mla6 + Mla14, Mla13 + MlRu3, Mla7 + MlNo3, Mla10 + MlDu2, Mla13 + MlRu3, Mlo-5 and virulent on the line with gene Mla8 (Table 1). Frequency of isolates virulent to genes, Mlg + MlCP and Mla12 + MlEm2 was very low. In some populations (Saryagash, Tulkubas, Ordabasy) isolates virulent to genes Mla22 and Mlat were detected at frequency of 11.1-33.3%. The isolates virulent to these two genes occurred frequently in European, Australian [19] and South African populations [21]. It has been recently reported that line P12 [10] with gene Mla22 is the most useful differential cultivar, i.e. possesses the maximal pathogen differentiating ability. However, none of commercial barley varieties includes this gene [20]. In our experiments isolates virulent to genes Mla3, Mlg + MlCP, Mla12 + MlEm2, Mla22 and Mlat were found only in some areas of South Kazakhstan region where winter varieties of barley are cultivated. In total virulence diversity of Bgh pathotypes was much higher in populations on winter barley varieties versus spring barley. It has been shown previously that the virulent pathotype (Va22) to gene Mla22 occurs only on winter barley varieties in France and fails to infect spring  [28]. They are carriers of effective resistance genes to Kazakhstanean Bgh isolates. Besides experience from Europe suggests the best ways of achieving durable resistance is to use either mlo [29] or combinations of minor genes. The recessive resistance gene mlo has remained effective for more than 50 years and is the mainstay of mildew control in European winter barley plantings [5].
Minimal intrapopulation diversity judging on all statistical indices (H W , H G , D) was displayed by the population from Zhambyl region, and maximal one by the population from South Kazakhstan population ( Table 5). The results of the study show significant difference even within the limits of one host variety. Spring barley variety Baisheshek is a source of infection in some areas of South Kazakhstan and Zhambyl regions ( Table 2). Analysis of pathotype structure of the fungus isolated from this cultivar showed higher virulence diversity of South Kazakhstan isolates versus Zhambyl isolates. Since there is no information on resistance genes in commercial barley varieties it is impossible to assess reliably which genes could influence the results of the virulence analysis. Powdery mildew in Kazakhstan is mainly an infection that is introduced from other countries, so one of the factors causing this difference could be different ways of the colony migration. Primary infection could be brought into South Kazakhstan region by airflows from adjacent Central Asian countries (Uzbekistan, Tajikistan), and into Zhambyl region by airflows from Kyrgyzstan and China. Virulence of the pathogens from Zhambyl population actually does not differ from that of the isolates from Chinese population [22], and the structure of Bgh populations in Uzbekistan and Tajikistan is still not known. The greatest virulence polymorphism of isolates was observed on lines with genes Mlra, Mlk, MlLa and Mlh (from 10.0 to 78.6%).
The data of many studies show that the virulence of the barley powdery mildew pathogen differs greatly in different eco-geographical regions. For instance, Czech population demonstrates high phenotypical diversity and the lowest mean complexity of virulence versus Israel population [18]. Virulence of Chinese Bgh populations differs substantially from the same of European populations; the only similarity consists in virulence to genes Mla8 and Ml(Ch) [22]. It has been recently shown that the structure of Bgh populations in South Africa is very variable, contains unique virulence frequencies and associations and differs from populations in other parts of the globe [21]. Comparison of our own results with published data displays low similarity of populations in Kazakhstan to European, African, Australian and South-West Asian populations. Particularly, the virulence rate of Bgh populations in Kazakhstan is much lower than in Europe [19,20], South Africa [21] and South-West Asia [18]. The difference between Kazakhstanean and Chinese populations consists mainly in the fact that the pathogens  Tables 3). Frequency of pathotype 24 occurrence reliably differes from the same of other pathotypes (significant difference from P < 0.004 to <0.0001). Significant difference in occurrence frequency was not observed between other pathotypes (Р > 0,05) in the surveyed regions. Statistical analysis was performed using two-way ANOVA followed by Tukey's multiple comparisons test. P values <0.05 were considered significant  [22]. The Australian population also differs from population in Kazakhstan, because the population of Bgh in our country contains the isolates virulent to gene Mlra that is effective in Australia [5].

Location of pathogen populations
The areas under barley were surveyed in 11 districts of four regions of Kazakhstan and adjacent Omsk region (Russia) in 2015-2016 (Fig. 1). The pathogen was present only in seven districts of two regions (South-Kazakhstan and Zhambyl) of Kazakhstan ( Figs. 1 and 2). Diseased samples were thus collected only from two regions within the surveyed districts in the period of high disease pressure from June to July. The distances between the survey regions ranged from 276 to 1155 km (Fig. 1), and between the sampling sites from 60 to 188.4 km (Fig. 2).

Sampling populations and multiplication of inoculum
A total of 107 isolates were collected during the survey period with most of the isolates (57%) being collected in 2016 (  [5,21,22]. Conidia from each colony were shaken onto leaf segments 25 mm long, which were excised from the central part of healthy, fully expanded primary leaves of the line B-3213 with no resistance gene to powdery mildew [31]. Inoculated leaf segments were placed in Petri dishes with water agar, prepared as above, and incubated under similar conditions for 10-11 days.

Differential sets and inoculation of leaf segments
The set of differentials used in 2015-2016 ( Table 1) was comprised of barley cv. Pallas and 16 near-isogenic 'Pallas' lines containing different genes for resistance to powdery mildew [10]. The differential set was kindly provided by Professor Mogens Støvring Hovmøller and Dr. Chris Khadgi Sørensen, Department of Agroecology, Aarhus University, Denmark. About 50 seeds of each differential were sown in a pot filled with a gardening peat substrate and placed in a greenhouse under natural daylight for 12-14 days. Leaf segments 20 mm long were taken from the central part of healthy fully expanded primary leaves of each differential. Three leaf segments of each differential were placed with the adaxial side facing up in a 150 mm glass Petri dish on the above-mentioned water agar. Leaf segments of differentials were inoculated in a metal inoculation tower 415 mm high and 150 mm in diameter. For each isolate, a glass Petri dish with leaf segments from the differential set was placed at the bottom of the tower. Inoculum of each isolate collected from a leaf segment was shaken onto a square piece (40 × 40 mm) of black paper to estimate the number of conidia deposited, and blown through a hole of 15 mm diameter in the upper part of the inoculation tower. The dishes with inoculated leaf segments were incubated in a chamber at 18 ± 2°C under artificial light (cool-white fluorescent lamps providing 12 h light at 30 ± 5 μmol m −2 s −1 ) [21,22].

Virulence determination
Reaction type (RT) on each differential and Bgh isolate combination was scored 8 days after inoculation on a 0-4 scale [32]. This scoring scale was supplemented with RT 0(4) (i.e., RT 0 with a few RT 4 colonies) [33], which is characteristic for barley lines carrying the mlo resistance gene. Isolates that produced RT 4 or 3-4 were considered virulent on the corresponding resistance gene(s). Virulence frequencies found during 2 years using seven individual populations are presented in Table 1. All tests for virulence of the Kazakhstanean populations were carried out at the Research Institute for Biological Safety Problems (Republic of Kazakhstan).

Pathotype designation
The isolates were assigned numerical designations based on their virulence to matching resistance genes in six differentials (Nos. 1-6 in Table 1). The differential set was divided into two triplets and each of the two digits indicates virulence or avirulence on the three differentials of the respective triplet. If a virulence to the corresponding resistance gene is detected, the first differential line has the value 1 (2 0 ), the second line has the value 2 (2 1 ), and the third line has the value 4 (2 2 ). Therefore, each digit can have a value from 0 (no virulence on any of the three differential lines) up to 7 (=1 + 2 + 4, virulent on each of the three differential lines). The resulting number based on six differentials defines the virulence of the isolates and consequently their classification as pathotypes [34,35].