Mapping and candidate gene screening of tomato Cladosporium fulvum-resistant gene Cf-19, based on high-throughput sequencing technology

Background Tomato leaf mold is a common disease in tomato cultivation. This disease is caused by Cladosporium fulvum, which has many physiological races and differentiates rapidly. Cf genes confer resistance to C. fulvum, and the C. fulvum-tomato pathosystem is a model for the study of gene-for-gene interactions. Plants carrying the Cf-19 gene show effective resistance to C. fulvum in the field, and can be used in breeding and resistance mechanism studies as new resistant materials. In this study, we used F2 bulk specific-locus amplified fragment sequencing (SLAF-seq) and parental resequencing methods to locate and characterize the Cf-19 gene. Results A total of 4108 Diff_markers and three association regions were found in association analysis. A 2.14-Mb region containing seven Cf-type genes was identified in further analysis based on data from SLAF-seq and parental resequencing. Two candidate genes, Solyc01g006550.2.1 and Solyc01g005870.1.1, were screened out by quantitative real-time PCR (qRT-PCR) analysis. Sequence analysis showed that Solyc01g006550.2.1 (an allelic locus of Cf-0) in CGN18423 was a novel homologue of the Cladosporium resistance gene Cf-9 (Hcr9s) in the Cf-4/9 locus. The marker P7, which cosegregated with the resistant trait, was developed based on sequence mutation of the Solyc01g006550.2.1 locus in CGN18423. Conclusions The Cf-19 gene was mapped to the short arm of chromosome 1. The candidate genes Solyc01g006550.2.1 and Solyc01g005870.1.1 showed related amino acid sequence structures and expression patterns. Solyc01g006550.2.1 had a close evolutionary relationship with the functional Hcr9 members Cf-4 and Cf-9, and was very different from non-functional members. The results from this study will facilitate the breeding of cultivars carrying the Cf-19 gene and provide a basis for further gene cloning, resistance gene evolution and plant resistance mechanism studies. Electronic supplementary material The online version of this article (doi:10.1186/s12870-016-0737-0) contains supplementary material, which is available to authorized users.

Results: A total of 4108 Diff_markers and three association regions were found in association analysis. A 2.14-Mb region containing seven Cf-type genes was identified in further analysis based on data from SLAF-seq and parental resequencing. Two candidate genes, Solyc01g006550.2.1 and Solyc01g005870.1.1, were screened out by quantitative real-time PCR (qRT-PCR) analysis. Sequence analysis showed that Solyc01g006550.2.1 (an allelic locus of Cf-0) in CGN18423 was a novel homologue of the Cladosporium resistance gene Cf-9 (Hcr9s) in the Cf-4/9 locus. The marker P7, which cosegregated with the resistant trait, was developed based on sequence mutation of the Solyc01g006550.2.1 locus in CGN18423. Conclusions: The Cf-19 gene was mapped to the short arm of chromosome 1. The candidate genes Solyc01g006550.2.1 and Solyc01g005870.1.1 showed related amino acid sequence structures and expression patterns. Solyc01g006550.2.1 had a close evolutionary relationship with the functional Hcr9 members Cf-4 and Cf-9, and was very different from non-functional members. The results from this study will facilitate the breeding of cultivars carrying the Cf-19 gene and provide a basis for further gene cloning, resistance gene evolution and plant resistance mechanism studies.
Keywords: Cf-19 gene, Cf-4/9 locus, Solanum lycopersicum, Cladosporium fulvum, Tomato leaf mold, SLAF-seq Background Tomato leaf mold disease, caused by the biotrophic fungus Cladosporium fulvum, is a serious disease of Solanum lycopersicum (tomato) [1]. This disease can reduce both fruit yield and quality, and sometimes even kill tomato plants. In compatible interactions with susceptible tomato plants, fungal spores germinate on the abaxial surface of leaves and enter the leaf apoplast through stomata. Hyphae emerge through the stomata and continue to grow and ramify. Finally, the infected cells undergo necrosis [2]. Tomato Cf genes confer resistance to C. fulvum and mediate incompatible interactions between C. fulvum and tomato plants. In incompatible interactions, fungal hyphae are arrested in their development soon after penetration of the sub-stomatal cavity, and their growth is restricted to limited necrotic lesions [2]. Plant breeders introduced Cf resistance genes from wild Solanum species into cultivated tomato to control the disease many years ago [3,4], and this is still an efficient method in tomato cultivation today. A number of Cf genes have been introgressed for use in commercial breeding, but this artificial selection has created evolutionary pressure on C. fulvum. To date, many Cf genes have been overcome by C. fulvum carrying matching avirulence genes (Avr). A race that has evolved to overcome the resistance genes Cf-2, Cf-4, Cf-5, Cf-9, and Cf-11 was reported by Lindhout et al. [5]. In northeast China, the C. fulvum physiological races 1.2.3.4 and 1.2.4 have evolved to overcome the Cf-4 gene and become the dominant races in only nine years [6]. The number of Cf genes that can be used in breeding is decreasing, making it necessary to identify new resistance genes and breed them into tomato cultivars.
Cf genes, which encode predicted membrane-bound proteins with extracytoplasmic leucine-rich repeats (LRRs) [7], confer resistance to specific races of C. fulvum through recognition of Avr peptides secreted into the leaf apoplast during infection [8]. In the tomato-C. fulvum interaction, a strict correlation exists between the triggering of a hypersensitive response (HR) and resistance, as the various Avrs induce a specific HR in tomato genotypes carrying a matching Cf resistance gene [9]. In 1980, 24 Cf genes located on 12 chromosomes were reported in tomato [10,11]. Later studies showed that the Cf genes were organized into clusters of resistance gene homologues, which have been designated Hcr2s and Hcr9s for homologues of the Cladosporium resistance genes Cf-2 and Cf-9, respectively [12]. Hcr2 loci, including the Cf-2 and Cf-5 loci, have been mapped to the short arm of chromosome 6 [13]. Three Hcr2 homologues were found at the Cf-2 locus. The Cf-2.1 and Cf-2.2 genes are nearly identical, and are functional genes that confer resistance to strains of C. fulvum that carry the matching Avr2 gene. The other homologue (Hcr2-2A) is not functional [14]. At the Cf-5 locus, four homologues were found, among which Hcr2-5C is the corresponding functional gene Cf-5 [15]. The Cf-4 and Cf-9 loci, each comprising five Hcr9s, have been mapped to the short arm of chromosome 1 [16]. Hcr9-4D and Hcr9-9C are the functional genes Cf-4 and Cf-9, respectively [7,17]. Tomato plants carrying the Cf-19 gene have shown efficient resistance in the field and no infection has been reported for this gene to date. Cf-19 was assigned to the long arm of chromosome 2 in 1980 [11]. Our previously study showed that Cf-19 was a dominant gene [18] and induced a remarkable HR in tomato plants inoculated with C. fulvum physiological race 1.2.3.4, indicating that it is a functional member of the Cf gene family.
SLAF-seq, which is based on high-throughput sequencing, is a recently developed high-resolution strategy for large-scale de novo single nucleotide polymorphism (SNP) discovery and genotyping [19]. This method is relatively low-cost and efficient, and can be used for gene and quantitative trait locus (QTL) mapping. The efficiency of this approach was tested in rice and soybean, and it was successfully used to construct genetic maps for sesame (Sesamum indicum L.) [20]. We applied this approach to an F 2 population in combination with genome resequencing of the parents to map and characterize the Cf-19 gene rapidly.

Results
Disease severity ratings and genetic analysis of the Cf-19 gene  Table 1).

High-throughput sequencing analysis
A total of 20.15 Gb data including 99.75 M reads and 4.07 Gb data including 20.22 M reads were obtained from parental genome sequencing and F 2 bulk SLAFseq, respectively ( Table 2). From the 119,504 SLAF tags, 4108 Diff-markers were obtained. A distribution diagram of the markers on each chromosome was drawn according to the results of SLAF positioning on the genome (Fig. 1).

Association analysis and candidate gene screening
According to the results of ΔSNP index calculation, all Diff-Markers were distributed on chromosome 1. Regions with three or more consecutive Diff-markers were regarded as association regions. We found three association regions including 203 Diff-markers and 244 genes on chromosome 1 in SLAF-seq analysis (Table 3). To further narrow the mapping region, SNPs in the association regions from parental resequencing were analyzed in combination with the F 2 bulk SLAF-seq data. Forty-three SNPs showed the Moneymaker base type in the susceptible pool, and both parental base types in the resistant pool were screened out. Of the 43 SNPs, 34 were distributed in an approximately 2.14-Mb region in association region I, one was in association region II and eight were in association region III (Additional file 1: Table S1). Further analysis was carried out based on the results of association

Quantitative real-time PCR analysis
According to the results of qRT-PCR analysis, of the seven candidate genes, two (Solyc01g005870.1.1 and Solyc01g006550.2.1) showed expression patterns related to the resistance response process. As shown in Fig. 3, the Solyc01g006550.2.1 gene was expressed at a low level before inoculation, and increased slightly after inoculation. This expression level was maintained for about 5 days, and then increased rapidly at 7 days after inoculation (DAI) and kept increasing during the following days. The highest expression was at 21 DAI, which was 62-fold higher than the 0 DAI value. The expression level of the Solyc01g005870.1.1 gene was relatively lower at every stage compared with the corresponding level of Solyc01g006550.2.1, while the general expression patterns of the two genes were similar. All other five genes showed unrelated expression patterns during the whole penetration process.
Multiple DNA sequence alignment of the candidate genes and other Cf-4/9 locus genes showed that the Solyc01g006550.2.1 locus of CGN18423 had a close evolutionary relationship with the Cf-4 and Cf-9 genes. The Solyc01g005870.1.1 locus of CGN18423 showed a short evolutionary distance to the Cf-0 gene, and the level of sequence divergence was very low (Fig. 5).

Marker development and linkage analysis
A sequence characterized amplified region (SCAR) marker (forward primer: 5'-AGTGCAGAAATGGGTTGT

Discussion
The Cf-19 gene was mapped to the short arm of chromosome 1 Cf-19 was assigned to the long arm of chromosome 2 in 1980 [11], while this gene was mapped at the short arm of chromosome 1 in this study. Although some Diff_-Markers were found on the long arm of chromosome 2 according to the results of F 2 bulk SLAF-seq analysis, they could not form association regions in further association analysis. Most Diff_Markers were found on the short arm of chromosome 1, and these markers showed us three association regions in association analysis. The position information for these regions was very reasonable for the Cf genes were organized into clusters of resistance gene homologues and most Cf genes mapped to chromosome 1 or chromosome 6 so far. Three association regions for the Cf-19 gene were identified in F 2 bulk SLAF-seq analysis. This result was not as accurate as we expected. All association regions in this study were on chromosome 1, and this may have been the reason why the location result was not accurate. The arrangements of Cf genes on the short arm of chromosome 1 are very complex. Two loci (Cf-4 and Cf-9) of the clustered Hcr9 genes have been mapped to this region, and each comprises five Hcr9s [16]. Additionally, the Cf-1 gene [21] has been mapped and closely linked to Cf-4/9 [16]. Other studies have mapped the Cf-ECP1, Cf-ECP2, Cf-ECP3, Cf-ECP4 and Cf-ECP5 genes to the short arm of chromosome 1 [22][23][24][25]. Many homologous fragments produced by evolutionary events are present in these genes and intergenic regions. These homologous sequences may influence the reads mapping and SNP statistics.  The Cf-4 gene originating from Lycopersicon hirsutum and the Cf-9 gene originating from L. pimpinellifolium were mapped to the Cf-4/9 locus [13]. This locus is flanked by conserved lipoxygenase sequences and is very complex, while only a single Hcr9 gene has been found at the Cf-4/9 locus in the disease-susceptible cultivar of Lycopersicon esculentum Moneymaker [7]; this gene was designated Cf-0. The Solyc01g006550.2.1 gene locus in Moneymaker (Solyc01g006550.2.1-Moneymaker) is just the Cf-0 locus; therefore, our candidate gene Solyc 01g006550.2.1 of CGN18423 (Solyc01g006550.2.1-CGN18423) is an allele of Cf-0. Sequence analysis showed that a 60-bp insertion was present in the Nterminal coding region, and this insertion resulted in changes in the ORF. Blast search results indicated that no genes were completely homologous to the Solyc01g006550.2.1-CGN18423 gene, which suggests that the candidate gene Solyc01g006550.2.1-CGN18423 is a novel member of the Cf-4/9 locus.
Cf genes encode proteins with classical signal peptide domains in their N-termini, LRRs, and a transmembrane region in their C-termini. Cf-4 and Cf-9 have identical C-termini [14], while a significant degree of sequence divergence is found in their N-terminal portions. This difference between Cf-4 and Cf-9 produces their recognition specificity [7]. A similar result was found in this study; the C-terminal portion of the candidate gene Solyc01g006550.2.1-CGN18423 had high identity to Cf-4 and Cf-9, while the N-terminal portion contained a high degree of sequence divergence. This result indicates that Solyc01g006550.2.1-CGN18423 may have recognition specificity that has formed based on a similar evolutionary mechanism to Cf-4 and Cf-9. LRRs can form a βstrand/β-turn motif in which the hypervariable residues are solvent-exposed and potentially contribute to recognition specificity. All Hcr9s at the Cf-4/9 locus encode proteins with 27 LRRs with the exception of 4B, which contains 23 LRRs, and Cf-4, which comprises 25 LRRs [7]. The candidate gene Solyc01g006550.2.1-CGN18423 encoded 30 LRRs, which is different to all other Hcr9s, providing conditions for the formation of a specific motif for ligand recognition. Phylogenetic analysis showed that the candidate gene Solyc01g006550.2.1-CGN18423 had very short evolutionary distances to functional Hcr9s including Cf-4, Cf-9, Hcr9-9A and Hcr9-9B (adult plants carrying 9A and 9B showed a resistance response to C. fulvum); these genes were clustered in a branch with a high degree of evolution, and were distinct from other nonfunctional Hcr9s. This result suggested a high possibility that the candidate gene Solyc01g006550.2.1-CGN18423 was our target gene Cf-19. The marker P7 can be used in marker-assisted selection (MAS) breeding P7 is a codominant marker that co-segregates with the resistance trait. This marker was designed based on a 60-bp insertion in Solyc01g006550.2.1, making the products in CGN18423 and Moneymaker different in size. This marker was tested in F 2 plants and different F 3 lines. Only three F 2 plants showed an unexpected genotype. These three F 2 plants were susceptible in the inoculation test, but showed the F 1 genotype in the P7 test. Hammond-Kosack and Jones [2] suggested that the increased C. fulvum invasion of host tissue and the higher titer of intercellular fluid required to elicit a necrotic or chlorotic response in lines where a Cf gene was present in a heterozygous state indicates that Cf genes are incompletely dominant. The Cf-19 gene may also be slightly affected by incompletely dominant inheritance, leading to a higher disease severity score for some heterozygous plants, and these plants were divided into the susceptible bulk. Although not all samples gave consistent results in the inoculation and P7 tests, the veracity of P7 in genotype identification is sufficient for MAS breeding work.

The combination of SLAF-seq and parental resequencing is effective for gene mapping
The SLAF-seq method provides significant advantages for developing large numbers of trait-related markers and target gene mapping. Genome resequencing can provide more details about genome sequences than SLAF-seq, and this information is useful to improve the precision of gene mapping. In this study, three association regions totaling 13.57 Mb in size were obtained in F 2 bulk SLAF-seq analysis, and a 2.14-Mb region with seven candidate genes was finally identified based on further SNP analysis using data from parental resequencing. The results of qRT-PCR analysis showed the accuracy of the mapping region at the transcriptional level. The current study also indicates that the combination of F 2 bulk SLAF-seq and parental resequencing is a good choice for both relatively low cost and high efficiency gene mapping and trait-related gene screening.

Impact of the current work on plant breeding and resistance mechanism research
Breeding to obtain resistant cultivars is an efficient method for the control of leaf mold outbreaks. The differentiation of C. fulvum physiological races is very rapid, and new Cf resistance genes are of great significance to breeding. Cf-19 was located on the short arm of chromosome 1 and one candidate gene Solyc01g006550.2.1 was assigned to the Cf-4/9 locus, while another candidate gene Solyc01g005870.1.1 was mapped close to the Cf-4/9 locus, which indicates that the Cf-19 gene may have a direct evolutionary relationship with Cf-4 and Cf-9. The characteristics of Cf-19 gene introgression to other cultivars may be similar to those of Cf-4 and Cf-9 and we have lots of experience in breeding cultivars carrying the Cf-4 or Cf-9 gene. Our location results will help guide the breeding of cultivars carrying the Cf-19 gene. The Cf gene family is important for studying plant resistance (R) gene evolution and R/Avr interaction mechanisms [26][27][28]. The candidate gene Solyc01g006550.2.1 is a new member of the Hcr9s and has a close evolutionary relationship with Cf-4 and Cf-9. It provides us a new gene material for plant resistance mechanism and R gene evolution research. The results we obtained in the present study were based on genetic mapping, expression pattern analysis and sequence analysis, but to clone the Cf-19 gene correctly, we still need functional verification such as virus-induced gene silencing (VIGS) and transgenic experiments.

Conclusions
In this study, we used F 2 SLAF-seq and parental resequencing to locate the Cf-19 gene. A total of 4108 Diff-markers were obtained. Three association regions consisting of 203 Diff-markers and 244 genes were found on chromosome 1. A 2.14-Mb region with seven candidate genes on the short arm was obtained through SNP analysis, and the candidate genes Solyc01g0 06550.2.1 and Solyc01g005870.1.1 were identified from these seven genes by qRT-PCR analysis. The candidate gene Solyc01g006550.2.1 is a new member of the Hcr9s in the Cf-4/9 locus. A SCAR marker (P7) that was cosegregant with the resistance trait was developed that can be used in MAS breeding work. These results provide a basis for Cf-19 gene cloning and application of the Cf-19 gene in breeding.

Plant materials and nucleotide extraction
The resistant line S. lycopersicum CGN18423 containing the Cf-19 gene (kindly provided by the Institute of Vegetable and Flowers, Chinese Academy of Agricultural Sciences) was crossed with the susceptible line S. lycopersicum Moneymaker. The resulting F 1 plants were self-crossed and backcrossed with Moneymaker to obtain F 2 and BC 1 . Three F 2 plants with different genotypes were self-crossed to obtain F 3 lines. All plants were grown at the Horticultural Experimental Station of Northeast Agricultural University.
At the five-to six-leaf stage, the seedlings of CGN18423, Moneymaker, F 1 , F 2 , BC 1 and F 3 plants were inoculated with C. fulvum race 1.2.3.4. The plants were assessed for disease severity at 15 DAI. Inoculation and assessment of disease severity ratings were performed as described by Wang et al. [6]. The disease severity symptoms of the plants were converted to a disease score of 0-9 points-0 points: no visible signs of infection; 1 point: 1-mm-diameter white spots or necrotic spots on the upper sides of leaves; 3 points: 2 to 3-mm-diameter yellow spots on the upper sides of leaves, some white mold on the lower sides of leaves, no spores formed; 5 points: 5 to 8-mm-diameter yellow spots on the upper sides of leaves, abundant white mold on the lower sides of leaves, a few spores formed; 7 points: 5 to 8-mmdiameter yellow spots on the upper sides of leaves, some black mold, many spores on the lower sides of leaves, also some black mold and no spores on the upper sides of leaves; 9 points: masses of spores formed on both sides of the leaves. Plants with a disease score of 0 to 3 points were classified as resistant whereas those with a score of 5 to 9 points were classified as susceptible.
Based on the inoculation results, 50 resistant and 50 susceptible plants were selected from the F 2 generation as two bulks for bulked segregate analysis (BSA) [29]. The bulked DNA samples were prepared by mixing an equal ratio of DNA extracted from 200 mg leaf samples using the cetyl trimethyl ammonium bromide (CTAB) method [30] with some modification [6]. These DNA samples were used for SLAF-seq analysis. DNA from the parents (CGN18423 and Moneymaker) and F 3 lines was also prepared using the CTAB method for parental resequencing and marker testing, respectively.
Leaf samples of CGN18423 and Moneymaker were collected at 0, 4, 5, 7, 10, 13, 17 and 21 DAI for qRT-PCR analysis. Total RNA was extracted from leaf samples using a plant RNA mini kit (Watson, Beijing, China) according to the manufacturer's handbook. First-strand cDNA was synthesized using an M-MLVRTase cDNA synthesis kit (Takara, Dalian, China) according to the manufacturer's instructions.

SLAF-seq and association analysis
A pre-design SLAF experiment was performed as described by [31]. According to the pre-design scheme, purified DNA was digested into DNA fragments of 364-464 bp in size, using an appropriate restriction enzyme (RsaI). All subsequent SLAF-seq procedures were carried out referring to Sun et al. [19]. In the association analysis, P stands for the susceptible parent Moneymaker, M means the resistant parent CGN18423, aa represents the resistant pool and ab represents the susceptible pool. There are two genotypes for a single marker; aa1 and ab1 mean the depth of aa and ab samples with the first genotype, respectively, and aa2 and ab2 stand for the depth of aa and ab samples with the other genotype, respectively. These variables were used in the following calculations: aa index = aa1aa1/aa2; ab index = ab1ab1/ab2; ΔSNP index = aa indexab index. The resistance phenotype is dominant, and the ratios of the two genotypes that can be detected are the same in unrelated markers, so the ΔSNP index of unrelated markers is equal to 0. Only one genotype of the ab sample can be detected in related markers; we stipulated that ab index equaled 0 or 1 and aa index had a value of 0.66 or 0.33, so the ΔSNP index of related markers was equal to −0.66 or 0.66, respectively. We calculated the ΔSNP index value of all markers and fitted these data; 95 % of the markers had fitting values greater than 0.3868831, so we chose this value to screen regions related to our target trait. Markers with fitting values higher than 0.3868831 were marked as Diff_Markers. Parental genome sequencing and data analysis DNA from CGN18423 and Moneymaker was used in this part of analysis. All steps including sequencing, reads mapping, and analysis of SNP and insertion/deletion (InDel) polymorphisms were carried out according to Bai et al. [32]. The tomato genome sequence of Heinz 1706 [33] was used as the reference genome for reads mapping. SNPs in the association regions obtained from parental resequencing were analyzed in combination with those from F 2 bulk SLAF-seq. SNPs that showed the Moneymaker base type in the susceptible pool, and both parental base types in the resistant pool were screened out to further narrow the mapping regions.

Gene annotation and candidate gene screening
Genes in the final mapping regions were annotated according to the tomato genome annotations in the National Center for Biotechnology Information (NCBI) (http://www.ncbi.nlm.nih.gov/BLAST/) and SOL Genomics Network (SGN) (http://solgenomics.net/) websites. All genes with Cf-type characters in function (receptorlike) and structure (LRR domain) were selected as candidate genes.
Quantitative real-time PCR analysis of candidate genes qRT-PCR was carried out for all seven candidate genes. The primers (Table 4) for qRT-PCR were designed using the Primer 5.0 software. The qRT-PCR reaction was performed using an iQ5 system (Bio-Rad, USA). The reaction mixture contained 10 μL of 2× TransStart Top Green qPCR SuperMix (TransGen, China), 10 pM of each primer, 2 μL of cDNA templates (1:10 dilution), and sterile distilled water to make up a total volume of 20 μL. The thermal conditions (two-step method) were as of 95°C for 10 s, T m temperature for 30 s. To detect primer dimerization or other artifacts of amplification, a melting-curve analysis was performed immediately after completion of the RT-PCR (95°C for 15 s and 55°C for 15 s, followed by a slow increase of temperature by 0.5°C per cycle to 95°C with continuous measurement of fluorescence). The data were analyzed using the 2 -ΔΔCT method [34] with EFα1 as a reference gene for normalization [35].

Candidate loci sequencing and DNA sequence analysis
To obtain detailed DNA sequence information, primers ( Gene structure analysis was performed using the online tool SMART (http://smart.embl-heidelberg.de/). Other known Cf genes in the Cf-4/9 locus (GenBank: AY639604.1, AJ002235.1, AJ002236.1) were also analyzed in combination with the candidate genes in this study to reveal evolutionary relationships.

Marker developing and linkage analysis
Six markers including cleaved amplified polymorphic sequence (CAPS) markers and SCAR markers were designed based on sequence mutations in or near the candidate gene loci. Marker primers were designed using the Primer 5.0 software. The markers were first screened using CGN18423, Moneymaker and the F 1 plants. Linkage analysis was then performed for these markers in an F 2 population consisting of 345 plants. In this part, a special SCAR marker P7 was screened out. Therefore, three F 3 lines (F 3 -1, derived from an F 2 plant with the resistant homozygous genotype; F 3 -2, derived from an F 2 plant with the resistant heterozygous genotype; F 3 -3, derived from an F 2 plant with the susceptible homologous genotype; based on the P7 test), were used to further test the genotyping veracity of the marker P7.

Availability of supporting data
The data sets supporting the results of this article are included within the article and its additional files.

Additional file
Additional file 1: Table S1. Results from combined data analysis of parental resequencing and F 2 SLAF-seq. Forty-three SNPs that showed the Moneymaker base type in the susceptible pool, and both parental base types in the resistant pool were screened out. Of the 43 SNPs, 34 were distributed in an approximately 2.14-Mb region in association region I, one was in association region II and eight were in association region III. Seven Cf-type genes were identified in association region I. (DOCX 16 kb)

Competing interests
The authors declare that they have no competing interests.