Comparative virulence phenotypes and molecular genotypes of Puccinia striiformis f. sp. tritici, the wheat stripe rust pathogen in China and the United States

f u n g a l b i o l o g y 1 1 6 ( 2 0 1 2 ) 6 4 3 e6 5 3

journal homepage: www.elsevier.com/locate/funbio

Comparative virulence phenotypes and molecular genotypes of Puccinia striiformis f. sp. tritici, the wheat stripe rust pathogen in China and the United States Gangming ZHANa,b, Xianming CHENb,c,*, Zhensheng KANGa,*, Lili HUANGa, Meinan WANGa,b, Anmin WANb, Peng CHENGb, Shiqin CAOd, Shelin JINd a

State Key Laboratory of Crop Stress Biology for Arid Areas and College of Plant Protection, Northwest A&F University, Yangling, Shaanxi, China b Department of Plant Pathology, Washington State University, Pullman, WA 99164-6430, USA c US Department of Agriculture, Agricultural Research Service, Wheat Genetics, Quality, Physiology, and Disease Research Unit, Washington State University, Pullman, WA 99164-6430, USA d Institute of Plant Protection, Gansu Academy of Agricultural Sciences, Lanzhou, Gansu, China

article info

abstract

Article history:

Stripe rust (yellow rust) of wheat, caused by Puccinia striiformis f. sp. tritici, is one of the most

Received 31 October 2011

important diseases in both China and the United States. The Chinese and US populations

Received in revised form

of the stripe rust fungus were compared for their virulence phenotypes on wheat cultivars

26 February 2012

used to differentiate races of the pathogen in China and the US and molecular genotypes

Accepted 22 March 2012

using simple sequence repeat (SSR) markers. From 86 Chinese isolates, 54 races were iden-

Available online 2 April 2012

tified based on reactions on the 17 Chinese differentials and 52 races were identified based

Corresponding Editor:

on the 20 US differentials. The selected 51 US isolates, representing 50 races based on the

Brenda Diana Wingfield

US differentials, were identified as 41 races using the Chinese differentials. A total of 132 virulence phenotypes were identified from the 137 isolates based on reactions on both

Keywords:

Chinese and US differentials. None of the isolates from the two countries had identical

Phylogenetic relationship

virulence phenotypes on both sets of differentials. From the 137 isolates, SSR markers iden-

Puccinia striiformis f. sp. tritici

tified 102 genotypes, of which 71 from China and 31 from the US. The virulence data clus-

Stripe rust

tered the 137 isolates into 20 virulence groups (VGs) and the marker data clustered the

Yellow rust

isolates into seven molecular groups (MGs). Virulence and SSR data had a low (r ¼ 0.34), but significant (P ¼ 0.01) correlation. Principal component analyses using either the virulence data or the SSR data separated the isolates into three groups: group a consisting of only Chinese isolates, group b consisting of both Chinese and US isolates and group c consisting of mostly US isolates. A neighbour-joining tree generated using the molecular data suggested that the P. striiformis f. sp. tritici populations of China and the US in general evolved independently. Published by Elsevier Ltd on behalf of The British Mycological Society.

* Corresponding authors. Department of Plant Pathology, Washington State University, 361 Johnson Hall, Pullman, WA 99164-6430, USA. Tel.: þ1 509 335 8086; fax: þ1 509 335 9581. E-mail addresses: [email protected], [email protected] 1878-6146/$ e see front matter Published by Elsevier Ltd on behalf of The British Mycological Society. doi:10.1016/j.funbio.2012.03.004

644

Introduction Puccinia striiformis Westend. f. sp. tritici Eriks. (Pst) causes stripe rust (yellow rust), one of the most important diseases of wheat worldwide. The pathogen was first officially reported in the United States in 1915 (Carleton 1915) and has frequently caused epidemics of significant damage since late 1950s (Line 2002; Chen 2005). In China, stripe rust was recorded in the historical literature several hundred years ago and has been scientifically studied since 1940s (Fang 1944; Li & Zeng 2002). Several devastating epidemics that caused yield losses of more than two million tons have occurred in China since 1950s (Wan et al. 2007). Virulence characterization of Pst populations has been conducted intensively in both China and the US, as well as some other countries, as the stripe rust pathogen evolves into new races (pathotypes) that may circumvent resistance in wheat cultivars and lead to devastating epidemics (Li & Zeng 2002; Chen 2005). Early studies on virulence races found that the set of wheat cultivars used to differentiate Pst races in Europe were not suitable to characterize races in both China (Fang 1944; Li & Zeng 2002) and the US (Line 2002). Therefore, different sets of differentials have been used in the two countries to best characterize the Pst populations in these countries. So far, there have been 67 races described in China, of which 33 with significant frequencies are considered as official Chinese races named as CYR (Chinese Yellow Rust) races and 34 with low frequencies are called ‘pathotypes’ (Li & Zeng 2002; Wan et al. 2004, 2007; Chen et al. 2009b). In the US, more than 140 Pst races have been reported since 1960s (Line & Qayoum 1992; Chen 2005, 2007; Chen et al. 2010; Wan & Chen 2012). However, because the previous studies used different differentials, little is known about common and different virulences in the Pst populations of China and the US, the two biggest stripe rust epidemiological regions in terms of area that can be affected by the disease in individual countries. Molecular characterization of stripe rust samples has been conducted using various techniques in the US (Chen et al. 1993, 1995; Markell & Milus 2008) and China (Shan et al. 1998; Zheng et al. 2001; Mboup et al. 2009; Duan et al. 2010). These studies generally revealed high levels of diversity within each of the countries. Before the present study, no efforts had been taken to compare the Pst molecular genotypes in both countries. The objective of this study was to compare virulence phenotypes and molecular genotypes of Pst in China and US through identifying races on both US and Chinese sets of differential cultivars and using simple sequence repeat (SSR) markers. Such a comparison may shed light to evolutionary mechanisms of pathogen variation, identify common effective resistance genes and provide virulence information for developing better strategies for effective control of the disease on a large scale.

Materials and methods Pst isolates A total of 137 Pst isolates were used in this study, of which 86 were collected from China and 51 from the US (Supplement Table 1). With exceptions of a few experimental isolates

G. Zhan et al.

representing specific races, most Chinese isolates were collected from 2005 to 2007 in wheat fields in Gansu, Guizhou, Henan, Shaanxi, Shandong, Sichuan and Yunnan provinces, covering the major wheat growing areas. The 51 US isolates were selected to represent the major races detected from 1960s to 2007 with the majority of the isolates collected from 2005 to 2007. Urediniospores were increased on seedlings of Nugaines, a US wheat cultivar susceptible in seedling stage to all Pst isolates tested so far (Chen et al. 2010; Wan & Chen 2012), following the standard procedures as previously described (Chen & Line 1992a, 1992b; Chen et al. 2002). Increased urediniospores were dried and kept in a desiccator at 4  C for later use. Fresh urediniospores or those kept in the desiccator at 4  C for less than 2 months were used to test a total of 37 wheat cultivars, including 17 used to differentiate Pst races in China (Wan et al. 2004) and 20 in the US (Chen et al. 2002; 2010) (Table 1). The standard procedures for identifying Pst races in the US (Chen & Line 1992a; Chen et al. 2002, 2010) were followed. Briefly, seedlings at two-leaf stage were dust-inoculated with urediniospores mixed with talc at a ratio of 1:20. Inoculated plants were placed in a dew chamber for 24 h at 10  C without light and then moved to a growth chamber to grow at a diurnal temperature cycle gradually changing from 4  C at 2:00 am to 20  C at 2:00 pm with a 16 h light and 8 h dark cycle. To prevent cross contamination, plants inoculated with different isolates were separated by plastic booths. Infection types (ITs) were recorded 18e20 days after inoculation using the 0e9 scale (Line & Qayoum 1992; Chen et al. 2002). An isolate producing an IT of 0e5 on a differential cultivar was considered avirulent (A) and 6e9 virulent (V). In this study, each isolate was tested at least twice on both Chinese and US differentials under the greenhouse conditions (Chen et al. 2002, 2010), except the following two situations: (i) The US isolates were tested only once on the US differential set during this study as their virulence patterns were previously determined (Line & Qayoum 1992; Chen 2005; Chen et al. 2010). (ii) Some Chinese isolates representing predominant races were tested on the Chinese differentials only once if the data was identical to the virulence pattern previously determined in China.

DNA extraction DNA was extracted directly from urediniospores using the method described by Aljanabi & Martinez (1997) with modifications. About 20 mg of dried urediniospores was used to extract genomic DNA for each isolate. To a 1.5 ml centrifuge tube containing urediniospores, 20 mg sterilized silica sand and 400 ml DNA extraction buffer [0.4 M NaCl, 10 mM TriseHCl (pH 8.0), and 2 mM EDTA (pH 8.0)] were added. The mixture was shaken for 2 min in a mini bead beater, added with 40 ml 20 % SDS and 8 ml of 20 mg ml1 proteinase K, mixed gently, and then incubated at 65  C for 3 h. After adding 300 ml 6 M NaCl, the mixture was centrifuged for 30 min at 12 500 rpm at 4  C. The top aqueous phase was transferred to a clean tube, and 400 ml chloroform was added, mixed gently and then centrifuged for 20 min at 12 500 rpm. The top aqueous phase was transferred to a new tube and an equal volume (about 400 ml) of cold isopropanol was added. The

Comparative virulence phenotypes and molecular genotypes of Puccinia striiformis f. sp. tritici

645

Table 1 e Wheat genotypes that are used to differentiate races of Puccinia striiformis f. sp. tritici (Pst) in the US and China used to determine virulence patterns of Pst isolates from the US and China. US differential No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Chinese differential a

Name

Yr gene

Lemhi Chinese 166 Heines VII Moro Paha Druchamp AvSYr5NIL Produra Yamhill Stephens Lee Fielder Tyee Tres Hyak Express AvSYr8NIL AvSYr9NIL Clement Compair

Yr21 Yr1 Yr2,YrHVII Yr10,YrMor YrPa1,YrPa2,YrPa3 Yr3a,YrD,YrDru Yr5 YrPr1,YrPr2 Yr2,Yr4a,YrYam Yr3a,YrS,YrSte Yr7,Yr22,Yr23 Yr6,Yr20 YrTye YrTr1,YrTr2 Yr17, YrTye YrExp1, YrExp2 Yr8 Yr9 Yr9,YrCle Yr8,Yr19

No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Name

Yr genea

Trigo eureka Fulhard Lutescens 128 Mentana Virgilio Abbondanza Early premium Funo Danish 1 Jubilejina II Fengchan 3 Lovrin 13 Kangyin 655 Suwon 11 Zhong 4 Lovrin 10 Hybrid 46

Yr6 Unknown Unknown Unknown YrVir1, YrVir2 Unknown Unknown YrA, þ Yr3 YrJu1, YrJu2, YrJu3, YrJu4 Yr1 Yr9, þ Yr1, YrKy1, YrKy2 YrSu Unknown Yr9 Yr4b, YrH46

a Refer to Chen (2005) and Chen et al. (2010) for the US differential cultivars and their Yr genes and Chen et al. (2009b) and Wan et al. (2004) for the Chinese differentials and their Yr genes for resistance to stripe rust.

solution was kept at 20  C overnight, and centrifuged for 5 min at 12 500 rpm to precipitate nucleic acid. The pellet was washed with 75 % ethanol and then washed with 95 % ethanol. After drying the DNA for 30 min in an ultra-clean hood and 100 ml sterile ddH2O was added to the tube to dissolve the DNA. To the DNA solution, 1.0 ml of 20 mg ml1 RNase was added and incubated at 37  C for 30 min. DNA concentration was determined using a ND-1000 spectrophotometer (Bio-Rad, Hercules, CA, USA) and stored at 20  C. For PCR amplification, the stock DNA solution was diluted to 50 ng ml1 as working solution and kept at 4  C.

PCR amplification Each PCR reaction of 15 ml consisted of 3.0 ml of 5 Taq polymerase buffer (with a final concentration of 50 mm KCl, 0.1 % Triton X-100, 10 mm TriseHCl, pH 8.0); 1.5 ml of 2.5 mM of the dATP, dCTP, dGTP, and dTTP mixture (Sigma Chemical., St. Louise, MO); 1.5 ml of 25 mM MgCl2; 1.2 ml of 5.0 mM of each primer; 1.2 ml working solution DNA as template; 0.12 ml of 5 unit/ml Taq DNA polymerase (Promega, Madison, WI, USA); and 5.28 ml sterile ddH2O. After initial screening, 12 SSR primer pairs (Table 2) were selected, of which four (CPS01, CPS02,

Table 2 e SSR primers used to identify polymorphic markers among Puccinia striiformis f. sp. tritici isolates, their repeat motifs, primer sequences, fragment size, annealing temperature, and the number of alleles identified in this study. Locusa

CPS01 CPS02 CPS11 CPS13 RB10 RJ3 RJ12 RJ17 RJ18 RJ20 RJ21 RJ24

Primers (50 30 )

Repeat motif

(TC)13 (GGT)5 (CAG)14 (GAC)6 (GT)7þ4þ4 (TGG)8 (AC)7 (GTT)5þ(GTT)7 (TGT)5 (CAG)4 (GTT)6 (GTT)5þ9

Forward

Reverse

TTAGGAGTAGCCCATCATC GGAGGAAGGGAATCAGTTCG GATAAGAAACAAGGGACAGC TCCAGGCAGTAAATCAGA CGC TAAGATTGGTGGTATGTGGTGGA GCAGCCTGGCAGGTGG ATCATTCCGATTTCTTTCTCACC TGGTGAGTGATGAGCTGG CTGCCCATGCTCTTCGTC AGAAGATCGACGCACCCG TTCCTGGATTGAATTCGTCG TTGCTGAGTAGTTTGCGGTGAG

GCA TGA AAC GAT CAA AGA AG CGC AGA CAA CCA ACT ATC ACG CAG TGA ACC CAA TTA CTC AG ATC AGC AGG TGT AGC CCC ATC TTGTCTTTCATCTCATCCAGCC GAT GAA TCA GGA TGG CTC C TCA CAC TGA TCC CAA TAG ATC AG ACA GCA ACA AAC TCA CCC ATC GAT GAA GTG GGT GCT GCT G CCT CCG ATT GGC TTA GGC CAG TTC TCA CTC GGA CCC AG CTC AAG CCC ATC CTC CAA CC

Annealing temp. ( C)

Size of amplified bands

55 58 55 55 52 52 52 52 52 52 52 52

366, 372 109, 112 200, 203 127, 130 220, 224 208, 212 271, 273 278, 284 352, 358 294, 303 170, 176 268, 286

a The information of the CPS primers was from Chen et al. (2009b) and that of RB and RJ primers from Duan et al. (2003) Enjalbert et al. (2002).

646

G. Zhan et al.

CPS11, and CPS13) were developed from EST sequences (Chen et al. 2009a), and eight (RJ3, RB10, RJ12, RJ17, RJ18, RJ20, RJ21, and RJ24) from genomic sequences (Enjalbert et al. 2002; Duan et al. 2003). Primers were synthesized by Operon Biotechnologies, Inc. (Huntsville, AL, USA). PCR amplifications were carried out in a PE-9700 Thermal Cycler (Applied Biosystems, Carlsbad, CA, USA). The amplification program consisted of an initial denaturation at 94  C for 2 min; 35 cycles of 94  C for 30 s, 52e58  C (depending upon primers) for 30 s and 72  C for 30 s; and then at 72  C for 7 min. After amplification, 6 ml of formamide loading buffer [98 % formamide, 10 mM EDTA (pH 8.0), 0.5 % (W/V) xylene cyanol and 0.5 % (W/V) bromophenol blue] was added to the PCR product and the mixture solution was denatured at 94  C for 3 min, from which 4.0 ml was loaded on a 5 % denaturing polyacrylamide gel. Gel preparation, electrophoresis, and silverstaining were done following the procedure as described by Chen et al. (1998). To estimate the sizes of SSR markers, 3 ml of 0.5 mg ml1 DNA molecular weight marker pUC18/MspI (Sigma-RBI, St. Louis, MO, USA) was used in each gel.

Data analyses For cluster analysis to determine relationships among isolates, avirulent ITs (0e5) were converted to 0 and virulent ITs (6e9) to 1 following the method of Chen et al. (1993, 1995). The co-dominant SSR marker data were converted to ‘2 0’ for one allele, ‘2 0’ for another allele, and ‘1 1’ for both alleles present, representing the homokaryotic states of each allele and heterokaryotic alleles (Rohlf 2008). No null allele was observed for any of the marker loci in the tested isolates. A similarity matrix based on simple match was generated using the SIMQUAL program of the NTsyspc 2.21 program (Rohlf 2008). Cluster analysis was done with the Unweighted PairGroup Method with Arithmetic (UPGMA) in the SAHN program of the NTSYS program. The dendrogram with the best fit to the similarity matrix was chosen based on cophenetic values (COPH) and MXCOMP, a matrix comparison procedure of the NTSYS program. Bootstrap analysis was used to determine robustness of branches of the dendrograms with the Winboot program (Nelson et al. 1994). Principal component analysis in the NTSYS program was used to construct a threedimensional figure to show more detailed relationships of Pst isolates (Rohlf 2008). Correlation between the SSR and virulence data was determined by comparison of the two similarity matrices using MXCOMP (Rohlf 2008). To determine

the phylogenetic relationships among the Chinese and US isolates, a neighbour-joining tree was constructed based on Nei genetic distances using the molecular data using the NTSYS program (Rohlf 2008). To study population structures, Nei (1973) and Shannon information indices (I ) (Shannon & Weaver 1949) were used to calculate the population diversities. As these indices treat non-identical phenotypes as equally distinct and do not count for race similarities, we also used the Kosman index, which takes both phenotypic frequencies and degrees of similarity among distinct phenotypes into consideration (Kosman 1996). The Kosman index, Ko, of the given population was calculated using the equation: Ko ¼ Assmax (A,A)/nk, where Assmax (A,A) is the maximum value of the sum of distances between matched pairs of isolates, k is the number of differentiating factors, and n is the number of isolates. The gene diversity (Ht) in all tested isolates was calculated as the sum of average gene diversity among populations of the two countries (Dst) and the average gene diversity within populations (Hs). The genetic differentiation (Gst) among the populations was computed as Gst ¼ Dst/Ht. Gene flow was estimated from Fst, which is equivalent to Gst for multiple allelic loci, as Nm ¼ 0.25(1  Fst)/Fst (McCallum et al. 1999). A genetic distance between populations was estimated for virulence and molecular data using Nei’s unbiased genetic distance coefficient (Nei 1973). The mean genetic similarity (GS) between populations was obtained by averaging individual GS estimates using the whole set of isolates belonging to the regional populations being compared. All calculations and analyses were conducted using the software POPGENE version 1.31 (Yeh et al. 1997), except for Kosman index that was calculated using the Virulence Analysis Tools (VAT) program (http:// www.tau.ac.il/lifesci/departments/plant_s/members/kosman/VAT.html). Analysis of molecular genetic variance (AMOVA) was conducted to statistically assess genetic variations within and among the Chinese and US populations with the software package Arlequin version 3.11 (Schneider et al. 2000).

Results Races of Chinese isolates When tested on the 17 Chinese differentials, the 86 Chinese isolates were identified as 54 races (Table 3, Supplement Table 1). The spectrum of virulence factors (virulence on

Table 3 e Numbers of isolates of Puccinia striiformis f. sp. tritici (Pst) from China and the US used in this study, races based on different sets of wheat differentials, VGs, genotypes based on SSR markers, and MGs. Pst population

Chinese US Total/overall

No. of isolates

86 51 137

No. of races based on differential sets Chinese

US

Chinese and US

54 41 93

52 50 97

81 51 132

No. of VGsa

No. of SSR genotypes

No. of MGsb

12 10 20

64 28 92

4 5 7

a VGs, which were separated by the middle value (0.78) of the similarity range (0.56e1.00) of the cluster analysis. b MGs, which were separated by clusters in a neighbour-joining tree using the SSR marker data.

Comparative virulence phenotypes and molecular genotypes of Puccinia striiformis f. sp. tritici

individual differentials) ranged from 3 to 16. On the 20 US differentials, the 86 isolates were differentiated into 52 races. The spectrum of virulence factors ranged from 5 to 18. When all Chinese and US differentials were considered, 81 races were identified from the 86 isolates, almost every isolate representing a different race. The Chinese isolates were all avirulent on Chinese differential Zhong 4 and US differentials Moro (Yr10, YrMor) and AvSYr5NIL (Yr5), indicating that the resistance genes in these genotypes were effective against all tested isolates.

Races of the US isolates When tested on the 17 Chinese differentials, the 51 US isolates were differentiated into 41 races (Table 3). The spectrum of virulence factors ranged from 2 to 15. On the 20 US differentials, these isolates were differentiated into 50 races as they were selected to represent the major Pst races in the US. The spectrum of virulence factors ranged from 1 to 17. When all of the Chinese and US differentials were considered, each of the 51 isolates was a different race.

647

group a with only Chinese isolates, group b with both Chinese and US isolates and group c with only US isolates (Fig 2B). These three groups corresponded to the three groups separated by the virulence data (Fig 2A). Comparison of similarity matrices revealed a low (r ¼ 0.34), but significant (P ¼ 0.01) correlation between the SSR and virulence data. Phylogenetic relationships of the Chinese and US isolates are shown in a neighbour-joining tree generated with the SSR data (Fig 3). The 137 tested isolates were clustered into seven MGs, of which three (MG3, MG5, and MG6) had only one isolate, one (MG7) had two isolates, one (MG1) had 20, one (MG2) had 36, and one (MG4) had 76 isolates. The 20 isolates in MG1 were all from China. The majority of the isolates in MG2 were from the US (31 isolates) and only five Chinese isolates (C13, C38, C52, C74, and C75) were in this MG. MG4 was the largest group consisting of 60 Chinese and 16 US isolates. Except MG3 and MG5-MG7 which had only one or two isolates, MG1, MG2, and MG4 in Fig 3 generally corresponded to the a, c, and b groups in Fig 2B using the simple match similarity matrix (data not shown). In this group, Chinese and US isolates tended to form into different clusters. The results indicated that the two populations are different from, but also related to each other.

Virulence comparison of Chinese and US isolates None of the Chinese and US isolates shared an identical virulence pattern on both Chinese and US differentials. Only a few isolates had the same virulence factors in only one set of differentials, but not both. When the virulence data were used in cluster analysis, the Chinese and US isolates were classified into 20 VGs using the middle value (0.78) of the similarity range (Fig 1). Among the 20 VGs, seven had only one isolate each and 13 had 2e31 isolates. Of the seven single-isolate VGs, five (VG4, VG5, VG8, VG9, VG10) were Chinese isolates and two (VG15, VG17) were US isolates. Of the 13 multiisolate VGs, four (VG1, VG2, VG3, VG6) consisted of only Chinese isolates; five (VG12, VG14, VG16, VG19, VG20) consisted of only US isolates; and four (VG7, VG11, VG13, VG18) had isolates from both China and the US. The 20 VGs can be classified into three super groups (SGs). The SG a, including VGs 1e5, consisted of only Chinese isolates. The SG b, including VGs 6e14, consisted of both Chinese and US isolates that were generally separated into different VGs with exceptions of few isolates. The SG c, including VGs 15e20, consisted of isolates all from the US except one Chinese isolate (C38). The relationships of the isolates in the three super groups, generally corresponding to the a, b, and c groups, respectively, are best illustrated in Fig 2A, resulted from the principal component analysis. Although group b contained isolates from both China and the US, the US isolates in this group actually were further separated from the Chinese isolates by component 3. The results showed that although some similar virulence patterns were shared by the Chinese and US Pst populations, they were largely different.

Molecular characterization of Chinese and US isolates Using the 12 SSR markers, 92 molecular genotypes were identified, of which 64 were from the Chinese isolates and 28 from the US isolates (Supplement Table 1). The principal component analysis separated the molecular genotypes into three groups:

Genetic diversity, differentiation and gene flow To further compare the Chinese and US populations, different indices of genetic diversity were calculated for each population and the two populations together using both virulence and molecular data (Table 4). When measured with Nei, Shannon and Kosman indices, the Chinese and US populations in general had similar levels of diversity (Table 4). Although there were not any identical genotypes between Chinese and US isolates, the genetic differentiation between Chinese and US isolates was low (Dst ¼ 0.03), which was supported by the high value of gene flow (Nm ¼ 1.74) between the two populations (Table 5). The analysis of variance with the molecular data revealed 30 % (P < 0.0001) variation between the Chinese and US populations, while the majority (70 %, P < 0.0001) of the molecular variation was within the populations (Table 6).

Discussion Stripe rust affects most wheat growing areas and virulence phenotypes of the pathogen have been monitored for more than a half century in both China and the US (Chen 2005; Wan et al. 2007). Due to different wheat cultivars that are used to differentiate races of Pst in the two countries, it was almost impossible to compare virulences and races of the pathogen. In the present study, we tested isolates obtained from China and the US on both sets of wheat differentials and with the same molecular markers. Both virulence and molecular data allowed us to compare directly the Pst populations in both countries for revealing common and different virulences, genetic relationships and possible mechanisms of genetic variation. The Pst isolates from both Chinese and the US tested in this study have virulences to a large number of resistance genes. The Chinese isolates collectively have virulences to 16 of the 17 Chinese differentials and 18 of the 20 US differentials. Similarly, the tested US isolates have virulences to 15 of the 17

648

G. Zhan et al.

Fig 1 e Dendrogram showing the similarities of 86 Chinese isolates (with a C prefix) and 51 US isolates (with a U prefix) of Puccinia striiformis f. sp. tritici based on virulence patterns produced on 17 Chinese and 20 US differentials listed in Table 2 based on the unweighted pair group arithmetic mean. The number at each branch shows the bootstrap value with 2 000 replications.

Chinese differentials and 17 of the 20 US differentials. Both populations do not have virulence to AvSYr5NIL (Yr5), which is used in the US set of differentials (Chen 2005; Chen et al. 2010) and has been recently added to the Chinese set of

differentials (Chen et al. 2009b). The tested Chinese isolates do not have virulence to the US differential Moro (Yr10, YrMor), but this virulence is present in many US races (Chen & Line 1992a; Chen 2005; Chen et al. 2010; Wan & Chen 2012). The

Comparative virulence phenotypes and molecular genotypes of Puccinia striiformis f. sp. tritici

649

Fig 2 e Principal component analyses showing relationships of Chinese and US isolates of Puccinia striiformis f. sp. tritici. A: based on virulence data and B: based on SSR marker data. The three components are labelled as 1, 2, and 3 in each figure.

tested US isolates do not have virulences to Chinese differentials Zhong 4 and Hybrid 46, while the Hybrid 46-virulence is present in many Chinese isolates (Wan et al. 2004, 2007; Chen et al. 2009b). Although the Zhong 4-virulence was not detected in the Chinese isolates tested in the present study, it was recently found in China (Wang et al. 2010). Except for these differences, the Chinese and US populations share the majority of the virulences to the differentials in both differential sets. The stripe rust fungus has the ability to compile various virulences in individual isolates to form new races. In this study, eleven Chinese isolates were identified as Chinese race CYR32, which has the longest list of virulences (16 out of 17) based on the Chinese differentials and has been one of the most predominant races in China since 1994 (Wan et al. 2004, 2007; Chen et al. 2009b). Four of the 11 isolates also had the longest list of virulences (18 out of 20) on the US differentials, which is even longer than the virulences of US race PST-127 that is virulent on 17 of the 20 US differentials (Chen et al. 2010). In general, isolates that were virulent on more Chinese differentials were also virulent on more US differentials. However, some isolates had a long list of virulences on one set of differentials, but had a short list of virulences on the other set of differentials. For example, Chinese isolate C16 was virulent on only six of the 17 Chinese differentials, but virulent on 18 of the 20 US differentials (Supplement Table 1). Similarly, the US isolate U50 (race PST-127) was virulent on 17 of the 20 US differentials, but virulent on 11 of the 17 Chinese differentials. The Chinese and US isolates were compared for their diversity with both virulence and molecular data. Based on the virulence data, the Nei and Shannon information indices reveal a higher diversity in the Chinese population than the US, but the Kosman indices showed a higher diversity in the US population. As the Kosman indices took the similarities between isolates into consideration (Kosman 1996), the values of Kosman index are more informative. With the molecular data, the three indices indicate the same or similar levels of diversity. The

virulence data resulted in a relatively higher diversity than the molecular data for both the Chinese and US isolates, which may indicate a faster evolution of virulence than generally neutral SSR loci. In the present study, the US isolates were selected to represent major races while the Chinese isolates were mostly collected in 2005e2007 with only a few isolates to represent historical Chinese races due to the lack of isolates for most historical races. These samples are best for comparison of the recent Chinese population to virulences and races that have been identified in the US, but may not be adequate for comparison of the recent populations in the two countries, especially for genetic diversity measured by molecular markers. In the future, systemic collections of China and the US, and also from other countries, should be studied periodically to identify common and unique races and molecular genotypes, to compare population structures, and determine evolutionary mechanisms of the stripe rust pathogen. The relationship between the Chinese and US populations was determined by three analyses. Firstly, through comparing the individual populations within the two populations together, we can see that the diversity values of both virulence and molecular data of the combined populations were not much higher than those of the Chinese population for the Nei and Shannon indices, or the US population with the Kosman index. In contrast, when Kosman index was used to compare diversity of individual populations with the two populations in combination, a higher total diversity was identified for the two populations together than either of the individual populations using both data (Table 4). Secondly, the two populations were compared directly for gene diversity, differentiation, and gene flow (Table 5). The gene diversity between the populations (Dst) was very low and the differentiation between the populations (Gst) was also low. A high gene flow (Nm) value between population was detected with both virulence and molecular data. Thirdly, the AMOVA results revealed that the majority (70 %) of the total molecular variation observed in this study existed within populations and the

650

G. Zhan et al.

Fig 3 e Neighbour-joining tree showing phylogenetic relationships of 86 Chinese isolates (with a C prefix) and 51 US isolates (with a U prefix) of Puccinia striiformis f. sp. tritici based on Nei genetic distances using the SSR marker data. The number at each branch shows the bootstrap value in percentage with 2 000 replications.

variation between the populations (30 %) was much smaller (Table 6). All of these analyses indicate that the Chinese and US populations are different, but closely related to each other.

The MGs and clusters in the neighbour-joining tree (Fig 3) may indicate evolutionary relationships among the groups. In spite of the country of origin, the isolates tested in the present study were separated into two super MGs. The first super

Comparative virulence phenotypes and molecular genotypes of Puccinia striiformis f. sp. tritici

Table 4 e Nei’s gene diversity, Shannon’s information index and Kosman indices in the Chinese and US population of Puccinia striiformis f. sp. tritici. Population Virulence Chinese US Chinese and US SSR markers Chinese US Chinese and US

Table 6 e Analysis of molecular variance (AMOVA) between and within Chinese and US populations of Puccinia striiformis f. sp. tritici.

Ha

Ib

Koc

Source of variance

0.36 0.29 0.37

0.53 0.44 0.54

0.46 0.49 0.68

0.27 0.27 0.30

0.40 0.40 0.45

0.40 0.39 0.45

Between populations Within populations Total

a H ¼ Nei’s gene diversity (Nei 1973). b Shannon & Weaver 1949. c Ko ¼ Kosman diversity index (Kosman 1996).

group consisted of MG1 to MG3 and the second super group consisted of MG4 to MG7. Within each group, isolates from different countries tended to be separated into different MGs, except C13, C38, C52, C74, and C75 in MG2 which included mostly US isolates. Although MG4 included isolates from both China and the US, isolates from different countries tended form separate sub-clusters. These results indicate that both Chinese and US populations have two genetic backgrounds represented by the two super groups. The isolates in the same MGs but from different countries may have relatively close genetic relationships. Taking C38 for example, it was collected from Weiyuan, Gansu, China in 2007. Although its virulence pattern was very different from US isolates and other closely related Chinese isolates based on reactions on the US differentials, C38 had a virulence pattern (virulent on Chinese differentials 2, 3, 4, 5, 6, 7, 8, and 14) very similar to US isolates U13 (virulent on Chinese differentials 2, 3, 4, 5, 6, 7, 8, and 9) and U10 and U35 (both virulent on Chinese differentials 2, 3, 4, 5, 6, 7, 9, and 11), as well as to Chinese isolates C7 and C9 (both virulent on Chinese differentials 2, 3, 4, 5, 6, 7, 8, and 11) (Supplement Table 1). The US isolates that were closely related to the Chinese isolate C38 represent some old US races, such as PST-3 (U2) and PST-7 (U3). Such connection may indicate an early migration event, supporting the previous hypothesis that the US Pst population might be originally from Asia (Line 2002), which is supported by the virulence data discussed above. On the other hand, it is also possible that isolate C38 and other molecularly similar Chinese isolates might have been introduced from the US. Nevertheless, no matter of the origin, such an introduction should have occurred quite long ago as PST-3 and PST-7 were first identified in the US in 1964 and 1974, respectively (Line & Qayoum 1992) and hardly detected since 2000 (Chen 2005; Chen et al. 2010; Wan & Chen 2012).

Table 5 e Genetic differentiation and gene flow between the Chinese and US populations of Puccinia striiformis f. sp. tritici. Between the Chinese and US populations

Dst

Gst

Nm

Based on virulent analysis Based on SSR analysis

0.03 0.03

0.13 0.13

1.63 1.71

651

DF

SS

PV (%)

Variance components

P

1

149.44

30.02

2.27

<0.0001

135

714.49

69.98

5.29

<0.0001

136

863.93

100

7.56

DF ¼ degree of freedom. SS ¼ sum of squared deviation. PV ¼ percentage of variance.

The molecular data may reveal direct evolutionary relationships among some of races within each group. A total of 16 groups had two or more isolates with identical or nearly identical molecular genotypes (Fig 3). If we assume that a late race has evolved from an early race with an identical molecular genotype, for instance, US race PST-37 (U14 in MG2) may be evolved from race PST-1 (U1); PST-29 (U12 in MG2) from PST-3 (U2); PST-80 (U32 in MG4), PST-97 (U38), PST-98 (U39), PST-117 (U48), and PST-122 (U49) from PST-58 (U18); and PST-114 (U45 in MG4), PST-115 (U46), and PST-116 (U47) from PST-101 (U41) (Fig 3; Supplement Table 1). The genetic relationships support the previous hypotheses of evolutionary relationships among these US races (Line & Qayoum 1992; Chen et al. 2010). However, we could not determine such evolutionary relationships for the Chinese isolates with an identical molecular genotype, as they were collected in a relative short period. Nevertheless, we can conclude that these isolates are genetically related and have a common origin. For example, Chinese isolates C5 (collected from Gansu in 2005) and C53 (collected from Henan in 2005) were of the same molecular genotype (GT2) in MG1, and also had similar virulence patterns. Similar cases can also be found between isolates C18 and C42 (both collected from Gansu in 2007) with the same GT62 in MG4 and between isolates C58 (collected in 2001) and C62 (collected in 2006) from Sichuan with the same GT59 in MG4 (Fig 3, Supplement Table 1). In contrast, isolates with an identical molecular genotype may differ greatly in their virulence pattern. For example, isolates C40 (collected from Gansu in 2007) and C83 (collected from Yunnan in 2005) had an identical molecular type (GT54), but had very different virulence patterns (VG3 vs. VG1) (Fig 3; Supplement Table 1). Such contrasting molecular and virulence relationships were also evident among isolates C6 (Gansu, 2005), C12 (Gansu, 2007), C51 (Guizhou, 2006), C66 (Sichuan, 2005), C77 (Shaanxi, 1997), C81 (Yunnan 2005), and C82 (Yunnan, 2005); and among isolates C33 (Gansu, 2007), C56 (Sichuan, 2001), C60 (Sichuan, 2007), and C78 (Shaanxi, 2007) in MG4 (Fig 3, Supplement Table 1). The isolates with an identical molecular genotype, but different virulence patterns, may have been evolved through a divergent process, as reported by Chen et al. (1993) in the US Pst population. The isolates from different epidemiological regions sharing the same genetic background also provide evidence for genetic exchanges between populations in the southwestern (Yunnan, Guizhou and Sichuan) and northwestern epidemic

652

regions (Gansu) in China. All these results demonstrate that diverse races of Pst can have a similar genetic background and that molecular genotypes are more powerful in determination of genetic and evolutionary relationships among isolates or populations than just virulences. This study is the first to directly compare the Chinese and US Pst isolates for their virulences with both sets of differentials. Both differential sets produced similar numbers of races (92 with the Chinese differentials and 97 with the US differentials). As the US isolates were selected based upon their specific races, the number of races identified based on both sets of differentials were as expected. However, a large number of races were identified in the Chinese isolates. Even with the Chinese differentials, from the 86 isolates we identified 54 races, of which 46 were not previously described. Such high number of races is somewhat in contrary to the total of 67 races (33 ‘CYR’ races and 34 called ‘pathotypes’) that have been named so far (Wan et al. 2004, 2007; Chen et al. 2009b). Naming Pst races in China has been conservative in numbers. Since 1994 when CYR32 was identified (38), only one new race, CYR33, has been named (Chen et al. 2009b). In contrast, a total of 89 races have been identified in the US with the US differentials from 1994 to 2007 (Chen 2005, 2007; Chen et al. 2010; Wan & Chen 2012). When tested on the US differentials, the 86 Chinese isolates were identified as 52 races (Table 1). The number is comparable to the number of races identified in the US in the similar period of time (Chen 2005; Chen et al. 2010). Some isolates that could be considered less virulent with one set of differentials appeared to be more virulent when tested with the other set of differentials. This is due to the fact that the Chinese and US differentials do not share many common genes for stripe rust resistance. With few exceptions, virulences on most of both Chinese and US differentials were found in both countries. Resistance gene Yr5 and the unidentified gene(s) in Zhong 4 were effective against all tested isolates from both China and the US. Hybrid 46 (Yr4b, YrH46) was resistant to all tested US isolates and Moro (Yr10, YrMor) was resistant to all Chinese isolates. These genes may still have some value in breeding for stripe rust resistant cultivars, but a large number of genes in the other differentials are no longer very useful as virulences are common in both countries. As rust migrations have occurred between countries and between continents (Stubbs 1985; Nararajan & Singh 1990; Wellings & McIntosh 1990; Brown & Hovmøller 2002; Wellings et al. 2003; Hovmøller et al. 2008), new virulences or races appearing in one country can spread to another country. Therefore, breeding programs should pay more attention to utilizing non-race specific resistance and should avoid using single genes that are just locally effective.

Acknowledgements This research was supported by the US Department of Agriculture, Agricultural Research Service (Project No. 5348-22000014-00D) and Washington Wheat Commission (Project No. 13C-3061-3923). PPNS No. 0564, Department of Plant Pathology, College of Agricultural, Human, and Natural Resource Sciences, Agricultural Research Center, Project Number

G. Zhan et al.

WNP00823, Washington State University, Pullman, WA 99164-6430, USA and special fund for Agro-scientific Research in the Public Interest of China (200903035-02) and the 111 Project from the Ministry of Education of China (B07049). We thank Dr Weidong Chen and Dr Jayaveeramuthu Nirmala for critical review of the manuscript.

Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.funbio.2012.03.004.

references

Aljanabi SM, Martinez I, 1997. Universal and rapid salt-extraction of high quality genomic DNA for PCR-based techniques. Nucleic Acids Research 25: 4692e4693. Brown JKM, Hovmøller MS, 2002. Aerial dispersal of pathogens on the global and continental scales and its impact on plant disease. Science 297: 537e541. Carleton MA, 1915. A serious new wheat rust in this country. Science 42: 58e59. Chen CQ, Zheng WM, Huang LL, Lu NH, Kang ZS, 2009a. Isolation of microsatellite loci from expressed sequence tags (ESTs) library of Puccinia striiformis f. sp. tritici. Molecular Ecology Resources 9: 236e238. Chen WQ, Wu LR, Liu TG, Xu SC, Jin SL, Peng YL, Wang BT, 2009b. Race dynamics, diversity, and virulence evolution in Puccinia striiformis f. sp. tritici, the causal agent of wheat stripe rust in China from 2003 to 2007. Plant Disease 93: 1093e1101. Chen XM, 2005. Epidemiology and control of stripe rust [Puccinia striiformis f. sp. tritici] on wheat. Canadian Journal Plant Pathology 27: 314e337. Chen XM, 2007. Challenges and solutions for stripe rust control in the United States. Australian Journal of Agricultural Research 58: 648e655. Chen XM, Line RF, 1992a. Inheritance of stripe rust resistance in wheat cultivars used to differentiate races of Puccinia striiformis in North America. Phytopathology 82: 633e637. Chen XM, Line RF, 1992b. Identification of stripe rust resistance genes in wheat cultivars used to differentiate North American races of Puccinia striiformis. Phytopathology 82: 1428e1434. Chen XM, Line RF, Leung H, 1993. Relationship between virulence variation and DNA polymorphism in Puccinia striiformis. Phytopathology 83: 1489e1497. Chen XM, Line RF, Leung H, 1995. Virulence and polymorphic DNA relationships of Puccinia striiformis f. sp. hordei to other rusts. Phytopathology 85: 1335e1342. Chen XM, Line RF, Leung H, 1998. Genome scanning for resistance-gene analogs in rice, barley and wheat by highresolution electrophoresis. Theoretical and Applied Genetics 97: 345e355. Chen XM, Moore M, Milus EA, Long DL, Line RF, Marshall D, Jackson L, 2002. Wheat stripe rust epidemics and races of Puccinia striiformis f. sp. tritici in the United States in 2000. Plant Disease 86: 39e46. Chen XM, Penman L, Wan AM, Cheng P, 2010. Virulence races of Puccinia striiformis f. sp. tritici in 2006 and 2007 and development of wheat stripe rust and distributions, dynamics, and evolutionary relationships of races from 2000 to 2007 in the United States. Canadian Journal of Plant Pathology 32: 315e323. Duan X, Enjalbert J, Vautrin D, Solignac M, Giraud T, 2003. Isolation of 12 microsatellite loci, using an enrichment protocol, in

Comparative virulence phenotypes and molecular genotypes of Puccinia striiformis f. sp. tritici

the phytopathogenic fungus Puccinia triticina. Molecular Ecology Notes 3: 65e67. Duan X, Tellier A, Wan A, Leconte M, de Vallavieille-Pope G, Enjalbert J, 2010. Puccinia striiformins f. sp. tritici presents high diversity and recombination in the over-summering zone of Gansu, China. Mycologia 102: 44e53. Enjalbert J, Duan X, Giraud T, Vautrin D, de Vallavieille-Pope C, Solignac M, 2002. Isolation of twelve microsatellite loci, using an enrichment protocol, in the phytopathogenic fungus Puccinia striiformis f. sp. tritici. Molecular Ecology Notes 2: 563e565. Fang CT, 1944. Physiological specialization of Puccinia glumarum Erikss. and Henn. in China. Phytopathology 34: 1020e1024. Hovmøller MS, Yahyaoui AH, Milus EA, Justesen AF, 2008. Rapid global spread of two aggressive strains of wheat rust fungus. Molecular Ecology 17: 3818e3826. Kosman E, 1996. Difference and diversity of plant pathogen populations: a new approach for measuring. Phytopathology 86: 1152e1155. Li ZQ, Zeng SM, 2002. Wheat Rusts in China. China Agricultural Press, Beijing China. Line RF, 2002. Stripe rust of wheat and barley in North America: a retrospective historical review. Annual Review Phytopathology 40: 75e118. Line RF, Qayoum A, 1992. Virulence, aggressiveness, evolution, and distribution of races of Puccinia striiformis (the cause of stripe rust of wheat) in North America, 1968e87. US Department of Agriculture, Agricultural Research Service Technical Bulletin 1788: 44. Markell SG, Milus EA, 2008. Emergence of a novel population of Puccinia striiformis f. sp. tritici in Eastern United States. Phytopathology 98: 632e639. Mboup M, Leconte M, Gautier A, Wan AM, Chen W, de Vallavieille-Pope C, Enjalbert J, 2009. Evidence of genetic recombination in wheat yellow rust populations of a Chinese oversummering area. Fungal Genetics and Biology 46: 299e307. McCallum BD, Roefls AP, Szabo LJ, Groth JV, 1999. Comparison of Puccinia graminis f. sp. tritici from South America and Europe. Plant Pathology 48: 574e581. Nararajan S, Singh DV, 1990. Long-distance dispersion of rust pathogens. Annual Review of Phytopathology 28: 139e153. Nei M, 1973. Analysis of gene diversity in subdivided populations. Proceedings of the National Academy of Sciences 70: 3321e3323.

653

Nelson RJ, Baraoidan MR, Vera Cruz CM, Yap IV, Leach JE, Mew TW, Leung H, 1994. Relationship between phylogeny and pathotype for the bacterial blight pathogen of rice. Applied and Environmental Microbiology 60: 3275e3283. Rohlf FJ, 2008. NTSYSpc: numerical taxonomy system, ver. 2.20. Exeter Publishing Ltd, Setauket, NY. Schneider S, Roessli D, Excoffier L, 2000. Arlequin Ver. 2000: a software for population genetics data analysis. Genetics and Biometry Laboratory, University of Geneva. Shan WX, Chen SY, Kang ZS, Wu LR, Li ZQ, 1998. Genetic diversity in Puccinia striiformis Westend. f. sp. tritici revealed by pathogen genome-specific repetitive sequence. Canadian Journal of Botany 76: 587e595. Shannon CE, Weaver W, 1949. The Mathematical Theory of Communication. University of Illinois Press, Urbana, USA. Stubbs RW, 1985. Stripe rust. In: Roelfs AP, Bushnell WR (eds), The Cereal Rusts. Disease, Distribution, Epidemiology and Controls, vol. 2. Academic Press, Orlando, pp. 61e101. Wan AM, Chen XM, 2012. Virulence, frequency, and distribution of races of Puccinia striiformis f. sp. tritici and P. striiformis f. sp. hordei identified in the United States in 2008 and 2009. Plant Disease 96: 67e74. Wan AM, Chen XM, He ZH, 2007. Wheat stripe rust in China. Australian Journal of Agricultural Research 6: 605e619. Wan AM, Zhao ZH, Chen XM, He ZH, Jin SL, Jia QZ, Yao G, Yang JX, Wang BT, Li GB, Bi YQ, Yuan ZY, 2004. Wheat stripe rust epidemics and virulence of Puccinia striiformis f. sp. tritici in China in 2002. Plant Disease 88: 896e904. Wang BT, Hu XP, Li Q, Li GB, Kang ZS, 2010. First report of new strains of Puccinia striiformis f. sp. tritici pathogenic to Zhong 4 (Trititrigia) in China. Phytopathology 100: S133. Wellings CR, McIntosh RA, 1990. Puccinia striiformis f. sp. tritici in Australasia: pathogenic changes during the first 10 years. Plant Pathology 39: 316e325. Wellings CR, Wright DG, Keiper F, Loughman R, 2003. First detection of wheat stripe rust in Western Australia: evidence for a foreign incursion. Australasian Journal Plant Pathology 32: 321e322. Yeh FC, Yang RC, Boyle T, 1997. POPGENE. CIFOR and University of Alberta, Canada, Version 1.21. Zheng WM, Liu F, Kang ZS, Chen SY, Li ZQ, Wu LR, 2001. AFLP fingerprinting of Chinese epidemic strains of Puccinia striiformis f. sp. tritici. Progress in Natural Science 11: 587e593.