Collinearity Analysis and High-Density Genetic Mapping of the Wheat Powdery Mildew Resistance Gene Pm40 in PI 672538

The wheat powdery mildew resistance gene Pm40, which is located on chromosomal arm 7BS, is effective against nearly all prevalent races of Blumeria graminis f. sp tritici (Bgt) in China and is carried by the common wheat germplasm PI 672538. A set of the F1, F2 and F2:3 populations from the cross of the resistant PI 672538 with the susceptible line L1034 were used to conduct genetic analysis of powdery mildew resistance and construct a high-density linkage map of the Pm40 gene. We constructed a high-density linkage genetic map with a total length of 6.18 cM and average spacing between markers of 0.48 cM.Pm40 is flanked by Xwmc335 and BF291338 at genetic distances of 0.58 cM and 0.26 cM, respectively, in deletion bin C-7BS-1-0.27. Comparative genomic analysis based on EST-STS markers established a high level of collinearity of the Pm40 genomic region with a 1.09-Mbp genomic region on Brachypodium chromosome 3, a 1.16-Mbp genomic region on rice chromosome 8, and a 1.62-Mbp genomic region on sorghum chromosome 7. We further anchored the Pm40 target intervals to the wheat genome sequence. A putative linear index of 85 wheat contigs containing 97 genes on 7BS was constructed. In total, 9 genes could be considered as candidates for the resistances to powdery mildew in the target genomic regions, which encoded proteins that were involved in the plant defense and response to pathogen attack. These results will facilitate the development of new markers for map-based cloning and marker-assisted selection of Pm40 in wheat breeding programs.


Introduction
Powdery mildew, which is caused by Blumeria graminis f. sp tritici (Bgt), is a globally destructive disease of common wheat (Triticum aestivum L.). This disease often causes large yield and mildew and is the most commonly employed powdery mildew resistance gene in Chinese breeding programs [19], but new isolates that are highly virulent to Pm21 have been reported [29,30]. However, Pm40 confers strong resistance to powdery mildew in the field of both Henna and Sichuan Provinces of China [21,25-26]), and this gene is effective against nearly all isolates collected from the main wheat growing regions of China [23]. Thus, Pm40 may be widely used in future Chinese breeding programs as an alternative to Pm21, necessitating the identification and prioritization of molecular markers closely associated with Pm40 by markerassisted selection is an important task.
Comparative genomics analysis is also useful for the development of new molecular markers linked to targeted genes. Comparative genomics analysis has been applied in hexaploid wheat, which has a large genome and numerous repetitive DNA sequences and lacks assembled reference genome sequences [31]. The ESTs can not only be developed into EST-sequence-tagged site (STS) markers and used to construct high-density maps but can also be used in comparative genomics analysis with the available genome sequences of rice (International Rice Genome Sequencing Project 2005), sorghum [32], and Brachypodium distachyon (The International Brachypodium Initiative 2010). The International Wheat Genome Sequencing Consortium (IWGSC) published a chromosome-based draft of common wheat genomic sequence that makes it possible to search the wheat genome of the regions containing Pm40 combined with comparative genomics analysis [33]. Via comparative genomics analyses, several disease resistance genes in wheat have been used in map-based cloning, such as Lr34/Yr18/Pm38 [34] and the stripe rust resistance gene Yr36 [35].
In this research, to achieve the eventual objective of marker-assisted selection and mapbased cloning, we studied the inheritance of Pm40, constructed a high-density genetic linkage map of Pm40, performed comparative genomics analysis of the regions of Pm40 in PI 672538, and obtained 9 candidate genes related to powdery mildew resistance.

Ethics Statement
All the field experiments were permitted by Sichuan Agricultural University (SICAU) and only tested in the experimental plots owned by SICAU. Collecting and inoculating Bgt races did not involve endangered or protected species.

Plant Materials
Two wheat lines, the powdery mildew resistant line PI 672538 [26] that carries the Pm40 gene without PmL962 [27] and the susceptible line L1034, were selected from the F 7 populations of a cross between the susceptible line MY11 and the resistant line YU25. Powdery mildew resistance in YU25 is putatively derived from Th. intermedium [36,37]. A set of 46 F 1 plants, 601 F 2 populations and 579 F 2:3 lines from a cross of PI 672538/L1034 were used to conduct genetic analysis of the response to powdery mildew and construct a high-density linkage map of the Pm40 gene.

Powdery Mildew Evaluations
The prevailing local isolate Bgt15, collected from Yaan City, Sichuan province, was used to inoculate the parents and the genetic populations by dusting conidia at a density of 100-140 conidia/mm 2 . The Bgt15 was avirulent on materials carrying the Pm40 gene and virulent on MY11 [22]. The wheat seedlings were sown in pots (3 cm diameter) in a growth chamber (Microclima MC1750E, Snijders Scientific, Tilburg, Holland) under controlled conditions with a 14-h light period at 22°C and a 10-h dark period at 18°C for the day/night cycle. Wheat plants were inoculated with Bgt15 at three-leaf stage seedlings. The same five-week-old seedlings were transplanted on the field to reassess the responses to powdery mildew. Twenty-five hybrid seedlings were planted in a randomized design in 2.5-m rows with 30-cm spacing. The infection types were classified using a rating scale of 0 to 4 [38]. The infection types produced on plants or lines were recorded 3 times over a one-week interval after inoculation.

DNA Extraction and Bulked Segregant Analysis
Genomic DNA was extracted from seedling leaves using a previously described CTAB protocol [39]. A mixture of equal amounts of bulked DNA from 10 homozygous resistant and 10 homozygous susceptible F 2 individuals (genotypes based on the reactions of the F 2:3 lines) was used for bulked segregant analysis (BSA) [40]. The polymorphic markers between the resistant and susceptible parents and the bulked DNA were chosen to genotype the F 2:3 lines to construct the linkage map of Pm40.

Polymerase Chain Reaction (PCR)
For the initial polymorphic marker survey, gwm [41] and wmc [42] SSR markers located on the wheat chromosome 7B according to a previously constructed consensus map [43] were selected and used in BSA to screen for markers linked to the resistance gene. PCR (25-μl volume) was performed in a PTC-200 thermocycler (MJ Research, Watertown, MA, USA). SSR analysis was performed following a previously described procedure [41] with minor modifications. Each PCR mixture contained each SSR primer at a concentration of 200 nmol/L, 0.2 mmol/L dNTPs, 1.5 mmol/L MgCl 2 , 1 unit of Taq polymerase, and 60 ng of template DNA. PCR was performed following a previously described program [23]. Then, 4 μL of each PCR product was mixed with 2 μL of loading buffer and loaded onto a 6% non-denaturing polyacrylamide gel for separation and visualization by silver staining [44].

Development of EST-STS Markers
To increase the marker density of the map, we chose other published SSR markers located on chromosome 7BS that co-segregated with the resistance locus in BSA, but other SSR markers were tested that were not linked with Pm40. Based on the published locations of the six linked SSR markers on wheat chromosome 7BS, the Pm40 genetic map region was located on the wheat C-7BS-1-0.27 bin map [45]. A total of 67 EST-STS markers were developed based on ESTs mapped on chromosome deletion 7BS bin 1-0.27 using the software primer3 [46]. These markers were employed to screen polymorphisms between the resistant and susceptible bulked DNA to construct a high-density genetic map and identify orthologous genomic regions.

Comparative Genomics Analysis
The polymorphic EST-STS markers between PI 672538 and L1034 as well as the resistant and susceptible bulked DNA were used to identify orthologous gene pairs. The markers linked with Pm40 and corresponding EST sequences were analyzed by BLASTn at a 10 −5 threshold probability against the genome sequence databases of Brachypodium distachyon, Oryza sativa japonica and Sorghum bicolor. After putative highly conserved gene pairs were obtained, we continued to perform tBLASTx to conduct comparative genomic analysis of the genomic regions on both sides of orthologous gene pairs among Brachypodium distachyon, Oryza sativa japonica and Sorghum bicolor. Genomic regions with s high level of collinearity were identified as orthologous genomic regions containing the Pm40 locus and polymorphic EST-STS markers. Furthermore, the sequences of those genes in three genomic regions were used as queries for BLASTn at a 10 −5 probability threshold and minimum of 100 bp match length against a chromosome-based draft of the wheat genomic sequence "Assembly_MIPSv2REF_Bgeno me_-cleaned_rep-masked" from IWGSC (http://www.wheatgenome.org/). Then, the contigs with the best hit were employed for searching the remaining non-homologous genes via IWGSC genomic annotation. Functional annotation of genes was performed using the software Blas-t2GO 3.30. Genomic locations were determined in silico using the software Circos 0.6.4.

Inheritance of Resistance to Powdery Mildew in PI 672538
The results of resistance identification confirmed that PI 672538 was resistant and L1034 was susceptible to powdery mildew at both the seedling stage and in adult plants (Fig 1)

Identification of Microsatellite Markers Linked with Pm40
The Pm40 gene was previously mapped to wheat chromosome 7BS. Thus, 87 published SSR markers mapped to wheat chromosome 7BS were chosen to map the Pm40 gene. A total of 9 (10.3%) of 87 microsatellite markers were polymorphic between PI 672538 and L1034. Of these, 6 markers, Xwmc364, Xwmc335, Xwmc476, Xgwm297, Xwmc662 and Xgwm43, were linked with Pm40 after genotyping the resistant and susceptible F 2 bulked DNA.

Identification of EST Markers and Construction of a Genetic Pm40 Linkage Map
Of 67 EST-STS markers developed from sequences mapped on the chromosome 7B deletion, 7 EST markers were polymorphic between the PI 672538 and L1034 as well as the resistant and susceptible bulked DNA. The sequences of the EST-STS markers linked with Pm40 are presented in Table 2. The linked EST-STS and previous SSR markers were used to genotype the F 2:3 populations. The relationship between the Pm40 gene and the marker genotypes is presented in S1 Table Each marker locus segregated in 1:2:1 or 3:1 ratios. A linkage map spanning chromosome arm 7BS was constructed ( Fig 2B). In total, 13 polymorphic markers and Pm40 were located on the genetic map. The map spans 6.18 cM with an average distance of 0.44 cM between markers. Pm40 is narrowly flanked by the markers Xwmc335 and BF291338 with distances of 0.58 and 0.26 cM located in deletion bin C-7BS-1-0.27 (Fig 2A).

Comparative Genomic Analysis and Identification of Candidate Genes
The sequences of 7 EST-STS polymorphic markers flanking Pm40 were used as queries to search for orthologs in rice, sorghum and Brachypodium genomic sequences. The 3 markers BE423064, BE446359, and BF291338 are homologous to Bradi3g40350, Bradi3g41590, Bra-di3g41750 of Brachypodium; Os08g0521400, Os08g0538300, Os08g0540100 of rice; and Sb07g025790 of sorghum (Table 3). Based on these 3 orthologous gene pairs, the orthologous genomic regions containing Pm40 and 3 markers in Brachypodium, rice, and sorghum were identified (S2 Table). The relationships between the 3 markers and their orthologous regions are presented in Fig 2B-2E. A high level of genomic collinearity was observed between rice sorghum and Brachypodium. Comparative genomic analysis established the collinearity of the  the orthologous gene order was highly conserved among these species (S1 Fig). As such, all of these genes were employed to search and sort the orthologs in wheat chromosome 7BS. Among this 2.92cM region of wheat, a total of 87 loci (76 IWGSC predicted genes and 11 unpredicted) distributed among 85 contigs were identified between markers BE423064 and BE446359 (S3 Table). Then, these 0.6-Mbp contigs putatively carrying Pm40 were employed to identify the genes that were not orthologous with the other 3 species. An additional 10 genes were identified in these contigs.
In total, 97 genes were assigned to this region ( Fig 2F). The putative functions of these genes are shown in S4 Table. Among them, 82 genes were annotated from the NCBI database through BLASTx search at a 10 −5 probability threshold. Sixteen out of 82 genes were referred to as ''hypothetical protein" or ''predicted protein". Nine genes showed the functional  annotations that were involved in the disease resistant mechanisms (Table 4). Three out of 9 genes domains include disease resistance proteins RGA1 and RPM1 respectively, which were both annotated as belonging to the Nucleotide Binding Site (NBS)-Leucine Rich Repeats (LRR) class. The remain 6 genes were involved in mechanisms of plant defense and response to environmental stresses such as wounding and pathogen attack, including one of hypothetical proteins related in defense response to fungus (GO: 0050832). These 9 genes could be considered as candidates for the resistances to powdery mildew in target genomic regions.

The Contribution of the EST-STS Markers to the Construction of the High-Density Genetic Map
To get the candidate resistance genes, constructing a high-density genetic map is essential. ESTs provide abundant information for gene expression profiling. The large number of localized ESTs allows EST-STS markers to be applied in the construction of a high-density genetic map [46]. In contrast to SSR markers, EST-STS markers reflect functional differences in genes and are useful for conducting comparative genomic analyses [48]. Pm40 was previously located by only five SSR markers encompassing a genetic distance of 10.9 cM using an F 2 population of 213 individuals [22]. In the present study, we confirmed that four of the five previously identified SSR markers, Xwmc335, Xgwm297, Xwmc364 and Xwmc476, were also linked with Pm40 in the mapping populations derived from the PI 672538/L1034 cross. Additionally, two additional linked SSRs markers and seven EST-STS markers were also located with Pm40 in this research using an F 2 population of 579 individuals (Fig 2). The average genetic distances between these markers and Pm40 ranged from 2.18cM to 0.48cM. Furthermore, the newly developed EST-STS markers BE446359, BF473824 and BF291338 are closer to Pm40 than the previous flanking marker, Xgwm297. Because EST-STS markers are typically located in conserved regions of expressed genes, the most closely linked marker BF473824, is more suitable for use in molecular marker-assisted breeding than SSR markers. The distance between the marker Xwmc335 and Pm40 increased from 0.2 cM to 0.58 cM [22] (Fig 2); the discrepancies in reported distances between a single marker and Pm40 may result from differences in the population sizes and genetic backgrounds of the materials studied. The population in the present study included 579 F 2 individuals, larger than the size of 213 used in previous studies; thus, a high-solution map was constructed to identify candidate resistance genes.

Identified the Candidate Genes of Powdery Mildew with the Methods of Comparative Analysis
Comparative genomics is a powerful method to study the species without reference assembled genome sequences. Genomic resources for wheat improvement have lagged behind other major crops, such as maize and rice. Draft genome sequences of the wheat A-genome progenitor Triticum urartu and the wheat D-genome progenitor Aegilops tauschii and a chromosomebased draft sequence of hexaploid bread wheat (Triticum aestivum) genome have been reported [33, 49,50]. However, sequence assembly and annotations of wheat are not complete. Thus, the EST-STS markers developed from wheat provide an excellent tool for comparative genomics analyses and homologous cloning. In the present work, three EST-STS markers, BE423064, BE446359, and BF291338, were used to identify the orthologous genomic regions containing Pm40 in Brachypodium, rice, and sorghum (Fig 2). As the first sequenced gramineous crop, rice has been successfully subjected to comparative genomics analyses. For example, the stripe rust resistance gene Yr36 was cloned by analyzing collinear regions in rice chromosome 2 [35]. However, gene rearrangement between the orthologous regions of wheat and rice was observed in the process of cloning the resistance genes Lr10, Lr21 and Pm3b [51][52][53]. In the Pm40 genomic regions, the synteny levels of the orthologs between rice and Brachypodium, rice and sorghum, Brachypodium and sorghum are 75.6%/60.8% (90 of 119/148), 79.8%/50.5% (95 of 119/ 188) and 83.1%/65.4% (123 of 148/188), respectively ( Table 3). The collinearity of Brachypodium between rice and sorghum is higher than that in the other pairs. Moreover, previous research has indicated higher collinearity between Brachypodium and wheat compared with wheat and rice and sorghum based on comparative genomics analysis of disease resistance gene regions [31, [54][55][56]. Thus, the collinear regions of Pm40 in Brachypodium are likely more favorable as a reference to determine the gene order in wheat. By comparing our results of the Pm40 target intervals to the 7BS Genome Zipper described by IWGSC [33], we found that the order of the wheat genes was similar to that of the anchored Brachypodium genes. Unfortunately, we couldn't anchor the marker BE423064 to the contig data from the IWGSC Genome Zipper, although a region of 84.95cM to 85.46cM on 7BS was similar to our final results for the collinear region. Therefore, it is difficult to directly use the data from the wheat Genome Zipper in our research, like a previous report regarding the positional isolation of powdery mildew QTLs in barley [57]. Successful fine mapping of quantitative trait loci by using the synteny-based or zipper-based markers combined with primarily genome zipper and population sequencing analysis has been reported [58,59]. Based on previous research, 97 gene were identified according to our synteny analysis with EST-STS markers. By comparison, only about 70 genes were anchored in this region by the IWGSC Genome Zipper analysis. Therefore, according to these EST-STS markers, we obtained a high-solution genomic data to provide more useful information for identifying the candidates of the powdery mildew gene Pm40. Inspection of all predicted proteins located in the target regions within Pm40 genes, we found 9 candidate genes encoded proteins that were involved in plant resistance (Table 4). Among them, 3 genes disease were identified as belonging to NBS-LRR class, which represents one of the major classes of resistance genes. Like most NBS-LRR resistance proteins, RGA1 and RPM1 guard the plant against pathogens via a direct or indirect protein-protein interaction [60,61]. Two candidates generally were described that defend from pathogen attacks via adjusting the oxidation-reduction reaction, including the genes encoding "glucan Peroxidase 55" and "L-ascorbate peroxidase proteins" [62,63]. Another two genes were annotated as "Calcium-transporting ATPase plasma membrane-type protein" and "probable glucan endo-1,3-beta-glucosidase A6", which may play an important role in the signaling networks of the pathogen effectors [64,65]. One additional candidate gene encode an equence-specific RNAbinding protein "Pumilio 5" that regulates translation and mRNA stability by binding the 3'-UTR of target mRNAs [66]. That triggers a unique defense system which affects pathogens replication. [67]. The last one is referred to as an uncharacterized protein "hypothetical protein F775_32194", which identified from Aegilops tauschii. It seems that may be involved in defense response to fungus (GO: 0050832) [49]. Apart from these, 30 out of 97 (31%) genes did not hit characterised proteins with predicted functions (S4 Table). These uncharacterised genes also will be concerned in the future work.

Polymerization and Application of Powdery Mildew Resistance Genes, including Pm40 Located on Chromosome 7B
The first powdery mildew resistance gene mapped on chromosome 7BS, Pm40, is dominant [22], whereas the second powdery mildew resistance gene mapped on 7BS, Pm47, is recessive [68]. In addition, Pm40 was putatively derived from Thinopyrum intermedium, whereas Pm47 was derived from wheat. Molecular linkage marker analysis revealed that the SSR markers linked to Pm40, which was physically mapped to bin C-7BS-1-0.27 near the centromere, are different from Pm47, which was mapped to bin 7BS-1-0.27-1.0. Based on pedigree, inheritance, molecular marker experiments, and genetic location, these data reveal large genetic differences between Pm40 and Pm47. Other resistant genes, including Pm5a-Pm5e [9,53,69], mlxbd [70], PmH [71], mljy, and mlsy [72], were also mapped to 7BL. This information indicates that we can develop stable, durable resistance against powdery mildew lines or cultivars by pyramiding the different Pm genes via chromosome recombination. Moreover, closely linked markers may accelerate the process of generating recombinant plants. The molecular markers closely linked with the Pm gene on 7B, such as EST-STS markers BF291338 closed linked with Pm40 and BE606897 closed linked with Pm47, may allow a recombinant chromosome with multiple resistance genes to be constructed more rapidly and efficiently [22,68].

Potential Role of PI 672538 in Marker-Assisted Selection for the Wheat Breeding Program
The resistant line PI 672538 used in the present study was selected from the progeny of MY11 and YU25 [26]. YU25 was derived from the cross between common wheat and Thinopyrum intermedium [21,36]. However, genomic in situ hybridization demonstrated that there is no alien chromosome fragment in PI 672538 [73]. In addition, in the present research, the responses of PI 672538/L1034 and crosses of the F 1 , F 2 , and F 2:3 populations to powdery mildew in different growth phases revealed PI 672538 resistance at both the seedling stage and in adult plants (Fig 1). PI 672538 was also simultaneously tested by artificial inoculation and natural inoculation in the field (Fig 1C and 1D). PI 672538 was not only resistant to the race Bgt15 but also exhibited resistant to various complex localized races. Extensive resistance identification and agronomic characterization of PI 672538 and L1034 have revealed that the effective resistance of PI 672538 to powdery mildew and stripe rust afforded is conferred by the wheat stripe rust resistance gene YrL693 [36,71] and Fusarium head blight resistance [26], respectively. Moreover, PI 672538 also exhibits favorable agronomic and morphological traits [26]. In summary, PI 672538 is an ideal material for resistance breeding. Therefore, the identification of EST-STS markers BF291338 closely linked to Pm40 will be beneficial for marker-assisted selection in the wheat breeding program. Comparative genomics analysis of the Pm40 region may aid further studies aimed at map-based cloning of resistance genes.