Abstract
Genetic resistance is a successful strategy for management of clubroot (Plasmodiophora brassicae) of brassica crops, but resistance can break down quickly. Identification of novel sources of resistance is especially important when new pathotypes arise. In the current study, the reaction of 177 accessions of Brassica napus to four new, virulent pathotypes of P. brassicae was assessed. Each accession was genotyped using genotyping by sequencing to identify and map novel sources of clubroot resistance using mixed linear model (MLM) analysis. The majority of accessions were highly susceptible (70–100 DSI), but a few accessions exhibited strong resistance (0–20 DSI) to pathotypes 5X (21 accessions), 3A (8), 2B (7), and 3D (15), based on the Canadian Clubroot Differential system. In total, 301,753 SNPs were mapped to 19 chromosomes. Population structure analysis indicated that the 177 accessions belong to seven major populations. SNPs were associated with resistance to each pathotype using MLM. In total, 13 important SNP loci were identified, with 9 SNPs mapped to the A-genome and 4 to the C-genome. The SNPs were associated with resistance to pathotypes 5X (2 SNPs), 3A (4), 2B (5) and 3D (6). A Blast search of 1.6 Mb upstream and downstream from each SNP identified 13 disease-resistance genes or domains. The distance between a SNP locus and the nearest resistance gene ranged from 0.04 to 0.74 Mb. The resistant lines and SNP markers identified in this study can be used to breed for resistance to the most prevalent new pathotypes of P. brassicae in Canada.
Similar content being viewed by others
Introduction
Canola (Brassica napus L.), also known as oilseed rape, is grown around the globe. It is the largest crop in Canada, seeded on 9.1 M ha each year with a farm gate value of $16.7 billion Cdn in 2019 (www.canolacouncil.org/news/canola-council-reports-on-2019-highlighting-strength-of-value-chain-partnership/). Demand for a healthy oil for human consumption, biofuel production, and use of canola meal as a high quality feed for livestock, have produced strong prices and a steady market for canola products. Clubroot, caused by Plasmodiophora brassicae Woronin, is an important disease of canola and other brassica crops worldwide1,2,3. Use of clubroot-resistant (CR) canola cultivars has been the most effective and widely used strategy for clubroot management in Canada4,5. Management of clubroot disease using major resistance genes has been effective but not durable. For example, the resistance available in the first generation of CR canola cultivars in Canada has broken down rapidly to reveal the presence of many new pathotypes of P. brassicae6,7, which complicates breeding for resistance.
Differential systems to identify pathotypes, which consist of selected cultivars with a consistent reaction to individual pathotypes, are available for P. brassicae. The systems of Somé et al.8 and the European Clubroot Differential (ECD)9 are widely used in Europe, but the Williams10 system was selected for recommendations to producers in Canada because of its simplicity and suitability for the initial situation in Canada8. These three systems have recently been replaced in Canada by the Canadian Clubroot Differential (CCD) set6, which was designed to differentiate among the many new pathotypes recently identified in Canada. Pathotype 5X is the first new pathotype identified to be virulent on the first generation of Canadian clubroot resistant cultivars; 3A and 3D are the most common new pathotypes; and 2B is one of the most virulent new pathotypes in western Canada.
Only a couple of major clubroot resistance genes and some quantitative trait loci (QTL)11,12,13 have been identified in B. napus (AC genome), but strong efforts have been made to identify novel sources of resistance to clubroot in other Brassica spp. and transfer them into canola. The majority of genes for clubroot resistance that have been identified are from B. rapa subsp. Rapifera9. The resistance in B. rapa has been utilized successfully in breeding for resistance to clubroot in B. napus5,7,8. More than 10 major clubroot resistance genes have been identified and mapped to chromosomes of B. rapa14,15,16,17,18,19,20,21,22,23,24,25,26,27,28. Also, six major clubroot resistance genes and at least 10 QTL have been mapped to B. oleracea (C genome)29,30,31,32,33.
Genome-wide association studies (GWAS) provide a quick and precise approach for linkage mapping in QTL detection studies. GWAS uses the linkage disequilibrium (LD) between alleles within diverse populations to detect potential association between markers and the traits of interest. In the first application of GWAS to identify resistance to clubroot, a 60 K SNP array was used to map 10 QTL in B. napus34. Transcriptome-based associated has also been used to identify several QTL for resistance to clubroot in B. napus35.
Next-generation sequencing can be used to identify thousands of single nucleotide polymorphism (SNP) markers and provide a dense cover of SNPs across the host genome, even in plant species with large genomes36. Genotyping by sequencing (GBS) provides a cost-effective approach for generating high density SNP panels in diverse accessions37, and GBS-association mapping has been used in many host–pathogen systems38,39. However, only a few genome-wide association mapping studies have been conducted in the B. napus–P. brassicae system34,35.
In the current study, a core collection of 177 accessions from a large germplasm collection of B. napus from around the world was selected to conduct GBS-based GWAS in the B. napus-P. brassicae system. The objective was to identify novel sources of resistance to clubroot from a core collection of B. napus accessions by (1) screening the collection under controlled conditions for reaction to four new pathotypes of P. brassicae identified on canola in Canada, (2) assessing the genetic diversity and structure of the core collection, and (3) conducting association mapping of resistance to the four pathotypes of P. brassicae.
Results
Evaluation of clubroot reaction
Initially, the reaction of 671 accessions to pathotype 5X (the first pathotype identified that was virulent on the first generation of CR canola cultivars in Canada) was assessed. Only 21 accessions were resistant, defined as a disease severity index (DSI) ≤ 20. After this initial assessment had been completed, more new pathotypes were identified. The 21 accessions resistant to pathotype 5X and an additional 156 accessions were tested for resistance to the new (to western Canada) pathotypes 3A, 2B and 3D. The majority of accessions were highly susceptible (70–100 DSI), but 8 accessions were resistant to 3A (0–16 DSI), 7 accessions to 2B (0–11 DSI) and 15 accessions to 3D (0–20 DSI) (Fig. 1, Supplemental Table S1).
Frequency distribution of accessions of Brassica napus plotted against clubroot severity (disease severity index, DSI) for pathotypes (a) 5X, (b) 2B, (c) 3A, and (d) 3D.
The resistant accessions were clustered into four groups based on their responses to combinations of pathotypes. Group 1 contained 3 accessions that were resistant to all four pathotypes. Group 2 comprised 3 accessions that were resistant to 3 of the 4 pathotypes. Group 3 had 2 accessions that were resistant to 2 out of 4 pathotypes, and Group 4 contained 25 accessions that were resistant to single pathotypes (Table 1). The majority of resistant accessions were winter oilseed rape lines of European origin. Phenotypic data were transformed using rank-based inverse normal transformation to make the DSI values fit a normal distribution, which is required for parametric model-based association analysis (Supplemental Fig. S1).
Sequence analysis and SNP discovery
GBS analysis of the 177 B. napus accessions generated ~ 1.2 billion total reads and ~ 633 million good barcoded reads, which were split into three FASTQ fills. On average, there were 3.3 M read counts per sample (range ~ 1.8 to 7.7 M) and 3.1 M read counts mapped (range 76–96%). Sequence tags from each file were captured and merged to produce a master tag file of 4,253,499 sequence tags. The tags were then aligned to B. napus reference genome v4.1 using the TASSEL-GBS pipeline. A total of 2,217,292 (52.1%) tags were uniquely aligned to the reference, 1,220,090 (28.7%) aligned to multiple positions and 816,117 (19.2%) were not aligned. Uniquely mapped tags were used to calculate the tag density distribution at each site in the B. napus genome and for SNP calling.
The raw sequence data for SNP calling were also analysed using the TASSEL-GBS pipeline. A total of 399,234 unfiltered SNPs and 355,680 filtered SNPs were called for the 177 accessions, with a mean individual depth of 8.5 ± 2.0 SD and mean site depth of 6.7 ± 11.4 SD. Of the 355,680 filtered SNPs, 301,753 SNPs (84.83%) were mapped to the 19 chromosomes of B. napus and were kept for further analyses. The remaining SNPs were randomly distributed without specific chromosome assignment, and so were discarded.
Variant analysis and annotation
There were slightly more SNPs in the C-genome (160,174 SNPs) than the A-genome (141,579 SNPs). Chromosome A03 had the highest number of SNPs within the A-genome, while C03 contained the highest number of SNPs in the C- genome (Table 2). The mean density was 2.12 SNP/Kb across the 19 chromosomes. In general, SNP density was higher in the C-genome (2.55 SNPs/Kb) than the A-genome (1.70 SNPs/Kb). C07 had the highest number of SNPs per Kb (2.88) and A10 had the lowest (1.43) (Table 2). The vast majority of SNPs were bi-allelic (90%), and only 10% were multi-allelic (Supplemental Fig. S2). There was a positive correlation (r = 0.80) between chromosome length and the number of SNPs, but only a weak correlation (r = 0.30) between the number of SNPs and the number of SNPs per Kb.
The SNPs were annotated using SnpEff software. About 37% of SNPs were annotated within coding regions, 22% within introns, 31% within promoter regions, 0.3% within splice sites, and 9.7% mapped to other genetic regions (Supplemental Fig. S3a). A more detailed SNP annotation was performed using the Variant Effect Predictor (Supplemental Fig. S3b). For SNPs within coding regions, 17% were non-synonymous, 18% were upstream-gene variants, 9% were downstream-gene variants, 23% were synonymous variants, 14% were intron variants, 15% intergenic variants, and 4% were located in the splice site regions and 5′ and 3′ UTRs (Supplemental Fig. S3b). Overall, more SNPs were annotated to the A-genome than to the C-genome (Supplemental Fig. S3c).
Genetic diversity and population structure
For genetic diversity analysis, the SNP markers were filtered based on a minor allele frequency (MAF) of 0.05 and minimum sample count of 80%, which resulted in 104,184 high quality SNPs. The mean MAF was the same for the A- and C-genomes (MAF = 0.14). Chromosome C01 had the highest MAF (0.16), followed by C03 and A07 (0.15), and lowest MAF (0.12) was in chromosomes A09 and C09 (Table 2). The mean marker heterozygosity (He) was 0.06 and the mean accession heterozygosity was 0.14. The average polymorphic information content (PIC) was the same for the A- and C-genomes (0.26). PIC was highest in chromosome C01 (0.27) and lowest (0.24) in A09 (Table 2). The ratio of transitions (changes from A <—> G and C <—> T) to transversions (changes from A <—> C, A <—> T, G <—> C or G <—> T) was 3: 2.
Population structure analysis indicated that the core collection was comprised of seven major clusters (Fig. 2, Supplemental Fig. S4). Population 1 contained 24 accessions (13.6%) representing all continents, but North American accessions were most frequently represented (10/24). Population 2 contained 14 accessions (7.9%), exclusively from Asia and Europe. Population 3 contained 36 accessions (20.3%) from all continents, but mainly from Europe. Population 4 contained 14 accessions (8.5%), mostly from Europe and Asia. Population 5 contained three accession (1.7%) from Europe and Asia. Population 6 contained 68 accessions (38.4%), mainly from Europe. Population 7 included 17 accessions (9.6%), almost exclusively from Europe (Fig. 2).
Population structure analysis of 177 accessions of Brassica napus based on model-based Bayesian clustering using STRUCTURE for K = 7 groups.
Linkage disequilibrium analysis
Linkage disequilibrium (LD) in the association panel was calculated using Pearson’s r2 statistic on pairwise combinations of SNPs present across the 19 chromosomes of B. napus (Supplemental Fig. S5). The average LD (r2) across the genome was 0.15. The mean LD was 0.10 in the A-genome and 0.19 in the C-genome. LD values ranged from 0.01 in A09 to 0.19 in C01 (Table 2). Across the genome, LD decayed very rapidly (r2 = 0.23) within 1.6 Mb (Supplemental Fig. S5).
Genome-wide association analysis
Genome-wide association analysis for clubroot severity was conducted using a general linear model (GLM) for naïve and P + Q analysis, and a mixed linear model (MLM) for P + K and P + Q + K. Naïve refers to genotypes and phenotypes only, Q is structure, and K is kinship. The quantile–quantile (Q–Q) plots from all models revealed that, save for significant SNPs, the distribution of observed – log10(p) was closest to the expected distribution in the P + Q + K compared to other models, so associations were identified using this model. A significance level of P ≤ 0.05/N (N: number of SNPs) based on the Bonferroni correction and a less stringent suggestive significance level at P ≤ 0.5/N were selected.
Association analysis detected 13 SNPs associated with resistance to the four P. brassicae pathotypes distributed on chromosomes A01, A03, A04, A05, C03, C07 and C09, with two SNPs (one on A05 and one on C07) associated with resistance to 5X, four SNPs to 3A (three on A03 and one on C03), three SNPs to 2B (on A01) and six SNPs (four on A01, one on C03 and one on C09) associated with resistance to 3D (Table 3, Fig. 3).
Manhattan plots of association analysis using mixed linear model (MLM) model P + K for pathotypes (a) 5X, (b) 2B, (c) 3A and (d) 3D. The horizontal line represents the threshold of significance (− log100.5/104,824 SNPs = 5.32).
Four significant SNPs (A01_19406654, A01_19406667, A01_19406668 and A01_19406672) on chromosome A01 were associated with resistance to pathotypes 2B and 3D. These four SNPs were in the same B. napus gene, BnaA01g27830D, which encoded glutamate decarboxylase 5. For pathotypes 3A, two significant SNPs (A03_50605605 and A03_50605321) on chromosome A03 were detected in the same gene, BnaA03g05600D, and SNP C03_327139017 on chromosome C03 was detected in gene BnaC03g07170D. Both genes on chromosomes A03 and C03 encode galactose oxidase/kelch repeat superfamily protein (Table 3).
Candidate resistance genes
A Blast search of 1.6 Mb upstream and downstream regions (distance based on LD) from significant SNPs was performed to identify any nucleotide-binding-site/leucine-rich-repeat (NBS-LRR) resistance protein encoding genes. For pathotype 5X, a coiled coil / nucleotide binding site / leucine-rich repeat (CC-NBS-LRR or CNL) family was identified at 0.53 Mb from the significant SNP, A05_101850460, on chromosome A05. For the SNP on C07, no NBS-LRR gene was found (Table 4). For pathotype 2B, no NBS-LRR gene was found in LD vicinity of the SNP on A04, while for SNPs on A01, a putative disease resistance protein, At3g15700, was detected at 0.16 Mb from the SNPs (Table 4). For pathotype 3A, no disease resistance genes were found adjacent to SNP A03_50588774 on A03, but a gene encoding an LRR domain was found at 0.08 Mb from the SNP. For SNP C03_327139017 on C03, an avirulence-induced gene (AIG1) family protein was detected at 0.04 Mb from the SNP. For pathotype 3D, a toll-interleukin-1 receptor/nucleotide binding site/leucine-rich repeat (TIR-NBS-LRR or TNL) gene was found at 0.07 Mb from the SNP, C03_356260031, on C03 (Table 4).
Discussion
In the current study, GWAS was used to identify and map new sources of resistance to four recently identified pathotypes (5X, 3A, 2B and 3D) of P. brassicae in 177 accessions of B. napus from around the world. The majority of the accessions were highly susceptible to all four pathotypes (80–100 DSI), but ~ 10% showed strong resistance (0–20 DSI). This indicated that sources of strong resistance to clubroot were much less common in B. napus than in B. rapa24,25. In total, 13 SNPs were associated with resistance, of which nine SNPs were on the A-genome, and four on the C-genome. Although the A-genome (from B. rapa) appeared to carry more QTL for clubroot resistance than the C-genome, this result indicated that the C-genome (from B. oleracea) could also be a potential source for clubroot resistance. Interestingly, four SNPs on A01 were common for pathotypes 2B and 3D, which may indicate that there is a common QTL for resistance to the two pathotypes. Moreover, SNPs located on A03 and C03 were associated with resistance to pathotype 3A. A Blast search determined that these SNPs were located very close to or within the same gene on both chromosomes. Association of SNPs with a single gene on separate chromosomes indicated the occurrence of inter-genome gene duplication, which is a common phenomenon in B. napus40,41,42.
QTL for clubroot resistance have been identified previously in B. napus, and several have been mapped to chromosomes A01, A02, A03, A08, C02, C03, C06, C07, and C0934,35. In those previous studies, the disease reaction to only a single strain of P. brassicae was assessed, and a SNP array or associative transcriptome was used rather than GBS, which might explain the difference in the number and identity of QTL detected. In a recent GWAS study, 45 QTLs were identified against 13 strains on chromosomes A01 to A10, and C02, C03 and C0543. In the current study, two pathotypes (3A and 2B) included in a recent study43 were analysed using a larger and potentially more diverse germplasm collection. The current study, however, identified different QTL from those in the previous study43. One possible explanation for the difference between the studies could be the analysis of DSI. A rank-based inverse normal transformation was used to make the DSI values nearly fit the normal distribution required for parametric model-based association analysis in the current study, while highly skewed DSIs were used in the previous study43. In addition, different sources of B. napus collections in the current study from the previous report43 could be an important factor contributing to the discrepancy. We conclude that all 13 QTL identified in the current study are likely to be novel because they were located at different physical locations on the chromosomes from the QTL identified previously43.
Extensive efforts have been made to map the genes for resistance to P. brassicae in Brassica species via bi-parental mapping approaches. In our previous studies, eight genes / QTL loci for resistance to pathotype 3H or 5X or both were mapped into chromosomes A02 (Rcr8), A03 (Rcr1, Rcr2, Rcr4 and Rcr5) and A08 (Rcr3, Rcr9 and Rcr9wa), and one gene into chromosome C07 (Rcr7)13,14,22,23,24,25. In the current study, 13 SNPs were identified from seven chromosomes (A01, A03, A04, A05, C03, C07 and C09) for resistance to four new pathotypes (2B, 3A, 3D and 5X) of P. brassica from Canada. None of the previously identified genes/QTL for resistance to Canadian pathotypes reside in chromosomes A01, A04, A05, C03 or C09. Although three SNPs were identified in chromosome A03 where Rcr1, Rcr2, Rcr4 and Rcr5 were located and one SNP in chromosome C07 where Rcr7 was located, these four SNPs were located in different regions from the identified genes. Three closely linked SNP A03_50588774, A03_50605321 and A03_50605605 associated with resistance to 3A were identified in the 2.5 Mb region of chromosome A03, while Rcr1, Rcr2, Rcr4 and Rcr5 spanned in the region of 23–25 Mb of chromosome A03 in the B. napus reference genome. Furthermore, SNP C07_531619426 associated with resistance to 5X was located at 17.9 Mb of chromosome C07. However, Rcr7 was identified in B. oleraces, in a region equivalent to the 25 Mb of chromosome C07 in the B. napus genome.
The majority of disease resistance genes identified in plants have been classified as TNL or CNL proteins, with ~ 70% of NBS-LRR genes in the Brassicaceae family being TNLs44,45,46,47. In the current study, five putative resistance genes were identified within LD distance of the SNPs associated with resistance to the four pathotypes. The resistance genes belonged to the TNL family (one gene) and the CNL family (one gene), as well as other types of resistance genes. The gene BnaA03g065730, which mapped close to SNP A03_50588774, contained the LRR protein domain that is an essential part of NBS-LRR proteins48. Moreover, the gene BnaC03g07250D that was located at 0.04 Mb from SNP C03_327139017 encoded an avirulence-induced gene (AIG1) family protein that, according to the gene-for-gene hypothesis by Flor49, should belong to a disease resistance protein-encoding gene family.
Population structure can have a significant impact on GWAS. Structure analysis using the entire SNP panel from GBS indicated that the core collection was comprised of seven sub-populations. However, a previous study using a SNP panel obtained from the Brassica 60 K Illumina Infinium SNP array identified only two major subpopulation50. Similarly, the difference in LD between studies at the chromosomal level as well as at the genome level appeared to be associated with the density of molecular markers. In the current study, LD extended further in the C-genome relative to the A-genome. LD decayed very rapidly (r2 = 0.23) within 1.6 Mb, while the range of previous reported LD decay values was affected by the diversity within the germplasm collection that was examined51,52,53.
In summary, the current study identified several accessions of B. napus with high levels of resistance to one or more or the four important new pathotypes of P. brassicae examined. Genome-wide association mapping analysis detected and mapped 13 SNP loci associated with resistance to the four pathotypes. This information will be used in subsequent genetic analysis of bi-parental populations to verify the SNPs and fine map the functional genes responsible for resistance to each pathotype, and also used for marker-assisted breeding of resistance to clubroot in canola.
Materials and methods
Plant and pathogen materials
A collection of 671 B. napus accessions were obtained from three gene banks; the Plant Genetic Resources of Canada, the Centre for Genetic Resources of the Netherlands, and Agricultural Research Service of the United of America. Self-pollination was performed for each line under greenhouse conditions to reduce the level of heterozygosity. The accessions were evaluated for resistance to a field collection of pathotype 5X (strain LG02) that had been characterized by Dr. S.E. Strelkov using the Canadian Clubroot Differential6 system. Also, selected lines were tested for resistance to field collections of pathotypes 3A (strain F.P. 3–14), 2B (F.P. 183–14) and 3D (F.P. 1–14) of P. brassicae (also provided by Dr. Strelkov) and then genotyped using a GBS platform.
A core group of 177 accessions from 32 countries was selected for GWAS. This group included accessions of oilseed rape (146 accessions), fodder rape (21), Swede rape (7), rutabaga (2) and turnip (1) from Europe (123 accessions), Asia (29), North America (20), Oceania (2), South America (1), Africa (1), and one accession of unknown origin (Supplemental Table S1). The growth habit was predominantly winter type (129), with some spring types (48) (Supplemental Table S1).
Seedlings for GBS analysis were grown to the 3–4 leaf stage in a growth chamber. A small (100 mg) sample of leaf tissue was collected from each accession, immediately frozen in liquid nitrogen, lyophilized in a freeze dryer for ~ 48 h and ground to a fine powder using a tissue lyser (Qiagen, Newtown City, USA).
Resting spores of pathotypes 5X, 3A, 2B, and 3D of P. brassicae were increased on susceptible canola and stored as frozen clubbed roots at − 20 °C until needed. Resting spores were extracted from the frozen clubs as described by Strelkov et al.54, and adjusted to a concentration of 1 × 107 resting spores mL−1. Spores of each pathotype were applied separately to the host entries.
Evaluation of clubroot reaction
The experiment was conducted in a growth chamber under the controlled environment following the method similar to that described by Suwabe et al.16. Seed of each host genotype was pre-germinated on moistened filter paper in a Petri dish. One-week-old seedlings of each host line × pathotype combination with 12 plants each were inoculated by dipping the entire root system in the resting spore suspension for 10 s. The inoculated seedlings were then immediately planted in 6 cm × 6 cm × 6 cm plastic pots filled with Sunshine LA4 potting mixture, with one seedling per pot. The pots were thoroughly watered and transferred to a greenhouse at 21 °C ± 2 °C with a 16 h photoperiod. The potting mixture was kept saturated with tap water at pH 6.5 for the first week after inoculation and then watered and fertilized as required.
Six weeks after inoculation, the seedlings were gently removed from the potting mix, the roots of each plant were washed with tap water, and each root was rated for clubroot symptom severity on a 0 to 3 scale, where: 0 = no clubs, 1 = a few small clubs on less than one-third of the roots, 2 = moderate clubs (small to medium-sized clubs on 1/3 to 2/3 of the roots), and 3 = severe clubs (medium to large-sized clubs on > 2/3 of the roots). A disease severity index (DSI) was then calculated using the formula of Horiuchi and Horias, modified by Strelkov et al.54.
where n is the number of plants in a class; N is the total number of plants in an experimental unit; and 0, 1, 2 and 3 are the symptom severity classes.
Assessment of the disease reaction of each accession with a resistant response (DSI < 20) in the initial test was repeated two more times. Each of these repetitions provided a similar result.
Sequence analysis and SNP discovery
The accession sequences were analyzed using genotyping by sequencing (GBS) in four major steps; DNA sample preparation, library construction, library sequencing and SNP calling. For sample preparation, DNA extraction was performed using the DNeasy 96 plant kit as per the manufacturer’s instruction (Qiagen). To reduce the genome complexity, DNA was digested with ApeKI, a methylation-sensitive restriction enzyme. For library construction, the fragments produced by digestion were directly ligated to enzyme-specific adapters followed by PCR amplification. For sequencing, the samples were divided into two pools of 96 samples each, and assessed in two runs of Illumina HiSeq 2500 (Illumina Inc., USA). DNA alignment was generated with BWA software version 0.7.8-r455. The GBS-TASSEL pipeline55 was used for SNP calling, and VCF and HapMap genotype files were generated. Initial SNP filtration was performed with the following settings: minor allele frequency (MAF) > 0.01 and missing data per site < 90%. Accessions with too much missing data were removed. Depth, missingness and heterozygosity were calculated using VCFtools V.0.1.1256. Genotyping and SNP calling were performed at the Genomic Diversity Facility, Cornell University (http://www.bio-tech. cornell.edu/brc /brc/ services). Beagle 5.157 was used to impute missing genotypes.
Variant analysis and annotation
Variants were annotated to regions of the B. napus reference genome v4.1 downloaded at https://wwwdev.genoscope.cns.fr/brassicanapus/data/ using SnpEff58, and Variant Effect Predictor (VEP)59, and variant locations were characterised as coding, intron, splice site, promoter and intergenic regions, intergenic, upstream, downstream, and synonymous.
Genetic diversity and population structure
Population-based genetic diversity, including allele frequencies, minor allele frequency (MAF), and average heterozygosity, were computed using TASSEL 5.2.18 software60. Polymorphic information content (PIC) values were calculated for SNP markers using the formula (PIC = 1 − (maf2 + (1-maf)2))-(2maf2(1-maf)2)), where maf = is the minor allele frequency. The ratio of transitions to transversions was calculated using the Kimura 2-parameter model in MEGA761.
Structure analysis of the accessions was conducted using fastSTRUCTURE v2.262. The admixture model and correlated allele frequency were applied with a burn-in period of 50,000 iterations and 100,000 replications of Markov Chain Monte Carlo (MCMC).
Linkage disequilibrium (LD) analysis
LD decay across the B. napus genome was measured and a correlation matrix of r2 values was computed between all pairs of polymorphic SNPs with MAF ≥ 5% using the GAPIT V3 package63.
Genome-wide association analysis
Clubroot severity data (disease severity index, DSI) were transformed using rank-based inverse normal transformation using the rntransform function in the GenABEL R Library64. Associations were analyzed for all SNP markers with MAF ≥ 5%, and evenly distributed (1 SNP per 100 Kb) using the following models: general linear model (GLM) for naïve and P + Q, and mixed linear models (MLM) for P + K and P + Q + K, where naïve refers to genotypes and phenotypes only, Q is structure and K is kinship. A kinship matrix of the accessions was calculated and principle components analysis was used to account for population structure and accession relatedness. The association analysis was done using TASSEL v.560.
Candidate resistance genes
Using Blast2Go software65, the sequence regions neighboring and within LD of the significant SNPs were searched for candidate genes encoding disease resistance proteins that were potentially responsible for the resistance to each pathotype of P. brassicae.
References
McNaughton, I.H. Swedes and rapes—Brassica napus (Cruciferae). In: Smartt J, Simmonds NW (eds.) Evolution of crop plants, 2nd ed. Longman Scientific & Technical, London, UK. 68–75 (1995).
Dixon, G. R. The occurrence and economic impact of Plasmodiophora brassicae and clubroot disease. J. Plant Growth Regul. 28, 194–202 (2009).
Karling, J. S. The Plasmodiophorales 2nd edn. (Hafner Publishing Company Inc, New York, 1968).
Voorrips, R. E. Plasmodiophora brassicae: aspects of pathogenesis and resistance in Brassica oleracea. Euphytica 83, 139–146 (1995).
Diederichsen, E., Frauen, M., Linders, E., Hatakeyama, K. & Hirai, M. Status and perspectives of clubroot resistance breeding in crucifer crops. J. Plant Growth Regul. 28, 265–281 (2009).
Strelkov, S. E. et al. Virulence and pathotype classification of Plasmodiophora brassicae populations collected from clubroot resistant canola (Brassica napus) in Canada. Can. J. Plant Pathol. 40, 284–298. https://doi.org/10.1080/07060661.2018.1459851 (2018).
Sedaghatkish, A., Gossen, B. D., Yu, F., Torkamaneh, D. & McDonald, M. R. Whole genome DNA similarity and population structure of Plasmodiophora brassicae strains in Canada. BMC Genom. 20, 744. https://doi.org/10.1186/s12864-019-6118-y (2019).
Somé, A. et al. Variation for virulence on Brassica napus L. amongst Plasmodiophora brassicae collections from France and derived single-spore isolates. Plant Pathol. 45, 432–439 (1996).
Buczacki, S. T. et al. Study of physiologic specialization in Plasmodiophora brassicae: proposals for attempted rationalization through an international approach. Trans. Br. Mycol. Soc. 65, 295–303 (1975).
Williams, P. H. A system for the determination of races of Plasmodiophora brassicae that infect cabbage and rutabaga. Phytopathology 56, 624–626 (1966).
Werner, S., Diederichsen, E., Frauen, M., Schondelmaier, J. & Jung, C. Genetic mapping of clubroot resistance genes in oilseed rape. Theor. Appl. Genet. 116, 363–372. https://doi.org/10.1007/s00122-007-0674-2 (2008).
Rahman, H., Shakir, A. & Hasan, M. J. Breeding for clubroot resistant spring canola (Brassica napus L.) for the Canadian prairies: Can the European winter canola cv. Mendel be used as a source of resistance?. Can. J. Plant Sci. 91, 447–458 (2011).
Piao, Z., Ramchiary, N. & Lim, Y. P. Genetics of clubroot resistance in Brassica species. J. Plant Growth Regul. 28, 252–264 (2009).
Diederichsen, E., Beckmann, J., Schondelmeier, J. & Dreyer, F. Genetics of clubroot resistance in Brassica napus ‘Mendel’. Acta Hort. 706, 307–311 (2006).
Huang, Z., Peng, G., Gossen, B. D. & Yu, F. Fine mapping of a clubroot resistance gene from turnip using SNP markers identified from bulked segregant RNA-Seq. Mol Breed. 39, 131. https://doi.org/10.1007/s11032-019-1038-8 (2019).
Yu, F. et al. Identification of SNP markers associated with clubroot resistance Gene Rcr1 through bulked segregant RNA sequencing. PLoS ONE 11, e0153218. https://doi.org/10.1371/journal.pone.0153218 (2016).
Yu, F., Zhang, X., Peng, G., Falk, K. C. & Strelkov, K. E. Genotyping-by-sequencing reveals three QTL for clubroot resistance to six pathotypes of Plasmodiophora brassicae in Brassica rapa. Sci. Rep. 7, 4516 (2017).
Sakamoto, K., Saito, A., Hayashida, N., Taguchi, G. & Matsumoto, E. Mapping of isolate-specific QTLs for clubroot resistance in Chinese cabbage (Brassica rapa L. ssp. pekinensis). Theor. Appl. Genet. 117, 759–767 (2008).
Suwabe, K. et al. Identification of two loci for resistance to clubroot (Plasmodiophora brassicae Woronin) in Brassica rapa L. Theor. Appl. Genet. 107, 997–1002 (2003).
Saito, M. et al. Fine mapping of the clubroot resistance gene Crr3 in Brassica rapa. Theor. Appl. Genet. 114, 81–91 (2006).
Hayashida, N. et al. Construction of a practical SCAR marker linked to clubroot resistance in Chinese cabbage, with intensive analysis of HC352b genes. Eng. Gakkai. Zasshi. 77, 150–154 (2008).
Kato, T., Hatakeyama, K., Fukino, N. & Matsumoto, S. Fine mapping of the clubroot resistance gene CRb and development of a useful selectable marker in Brassica rapa. Breed. Sci. 63, 116–124 (2013).
Piao, Z. Y., Deng, Y. Q., Choi, S. R., Park, Y. J. & Lim, Y. P. SCAR and CAPS mapping of CRb, a gene conferring resistance to Plasmodiophora brassicae in Chinese cabbage (Brassica rapa ssp. pekinensis). Theor. Appl. Genet. 108, 1458–1465 (2004).
Zhang, T. et al. Fine genetic and physical mapping of the CRb gene conferring resistance to clubroot disease in Brassica rapa. Mol. Breed. 34, 1173–1183 (2014).
Chu, M. et al. Fine mapping of Rcr1 and analyses of its effect on transcriptome patterns during infection by Plasmodiophora brassicae. BMC Genom. 15, z1166 (2014).
Huang, Z. et al. Fine mapping of a clubroot resistance gene in Chinese cabbage using SNP markers identified from bulked segregant RNA sequencing. Front. Plant Sci. 8, 1448. https://doi.org/10.3389/fpls.2017.01448 (2017).
Karim, M. M. et al. Two clubroot-resistance genes, Rcr3 and Rcr9wa, mapped in Brassica rapa using bulk segregant RNA sequencing. Int. J. Mol. Sci. 21, 5033 (2020).
Dakouri, A. et al. Analysis of genome-wide variants through bulked segregant RNA sequencing reveals a major gene for resistance to Plasmodiophora brassicae in Brassica oleracea. Sci. Rep. 8, 17657 (2018).
Rocherieux, J. et al. Isolate-specific and broad-spectrum QTLs are involved in the control of clubroot in Brassica oleracea. Theor. Appl. Genet. 108, 1555–1563. https://doi.org/10.1007/s00122-003-1580-x (2004).
Voorrips, R. E., Jongerius, M. C. & Kanne, H. J. Mapping of two genes for resistance to clubroot (Plasmodiophora brassicae) in a population of doubled haploid lines of Brassica oleraceae by means of RFLP and AFLP markers. Theor. Appl. Genet. 94, 75–82. https://doi.org/10.1007/s001220050384 (1997).
Grandclément, C. & Thomas, G. Detection and analysis of QTLs based on RAPD markers for polygenic resistance to Plasmodiophora brassicae Woronin in Brassica oleracea L. Theor. Appl. Genet. 93, 86–90. https://doi.org/10.1007/BF00225731 (1996).
Landry, B. S. et al. A genetic map for Brassica oleracea based on RFLP markers detected with expressed DNA sequences and mapping of resistance genes to race 2 of Plasmodiophora brassicae (Woronin). Genome 35, 409–420. https://doi.org/10.1139/g92-061 (1992).
Manzanares-Dauleux, M. J., Divaret, I., Baron, F. & Thomas, G. Evaluation of French Brassica oleracea landraces for resistance to Plasmodiophora brassicae. Euphytica 113, 211–218. https://doi.org/10.1023/A:1003997421340 (2000).
Li, L. et al. A genome-wide association study reveals new loci for resistance to clubroot disease in Brassica napus. Front. Plant Sci. 7, 1483. https://doi.org/10.3389/fpls.2016.01483 (2016).
Hejna, O. et al. Analysing the genetic architecture of clubroot resistance variation in Brassica napus by associative transcriptomics. Mol. Breed. 39, 112. https://doi.org/10.1007/s11032-019-1021-4 (2019).
Gupta, P. K., Rustgi, S. & Kulwal, P. L. Linkage disequilibrium and association studies in higher plants: present status and future prospects. Plant Mol. Biol. 57, 461–485 (2005).
Elshire, R. J. et al. A robust, simple genotyping-by-sequencing (GBS) approach for high diversity species. PLoS ONE 6, e19379 (2011).
Flint-Garcia, S. A. et al. Maize association population: a high-resolution platform for quantitative trait locus dissection. Plant J. 44, 1054–1064 (2005).
Kump, K. L. et al. Genome-wide association study of quantitative resistance to southern leaf blight in the maize nested association mapping population. Nat. Genet. 43, 163–168 (2011).
Fopa Fomeju, B., Falentin, C., Lassalle, G., Manzanares-Dauleux, M. J. & Delourme, R. Comparative genomic analysis of duplicated homoeologous regions involved in the resistance of Brassica napus to stem canker. Front. Plant Sci. 6, 772 (2015).
Parkin, I. A., Sharpe, A. G. & Lydiate, D. J. Patterns of genome duplication within the Brassica napus genome. Genome 46, 291–303. https://doi.org/10.1139/g03-006 (2003).
Fopa Fomeju, B., Falentin, C., Lassalle, G., Manzanares-Dauleux, M. J. & Delourme, R. Homoeologous duplicated regions are involved in quantitative resistance of Brassica napus to stem canker. BMC Genom. 15, 498. https://doi.org/10.1186/1471-2164-15-498 (2014).
Fredua-Agyeman, R., Yu, Z., Hwang, S. F. & Strelkov, S. E. Genome-wide mapping of loci associated with resistance to clubroot in Brassica napus ssp. napobrassica (rutabaga) accessions from Nordic countries. Front. Plant Sci. 11, 742 (2020).
Mun, J., Yu, H., Park, S. & Park, B. Genome-wide identification of NBS encoding resistance genes in Brassica rapa. Mol. Gen. Genom. 282, 617–631. https://doi.org/10.1007/s00438-009-0492-0 (2009).
Yu, J. et al. Genome-wide comparative analysis of NBS encoding genes between Brassica species and Arabidopsis thaliana. BMC Genom. 15, 3. https://doi.org/10.1186/1471-2164-15-3 (2014).
Zhang, Y. M. et al. Uncovering the dynamic evolution of nucleotide-binding site-leucine rich repeat (NBS-LRR) genes in Brassicaceae. J. Integr. Plant Biol. 58, 165–177. https://doi.org/10.1111/jipb.12365 (2016).
Leister, D. Tandem and segmental gene duplication and recombination in the evolution of plant disease resistance genes. Trends Gen 20, 116–122. https://doi.org/10.1016/j.tig.2004.01.007 (2004).
Padmanabhan, M., Cournoyer, P. & Dinesh-Kumar, S. P. The leucine-rich repeat domain in plant innate immunity: a wealth of possibilities. Cell Microbiol. 11, 191–198. https://doi.org/10.1111/j.1462-5822.2008.01260.x (2009).
Flor, H. H. Inheritance of pathogenicity in Melampsora lini. Phytopathology 32, 653–669 (1942).
Strelkov, S. E., Tewari, J. P. & Smith-Degenhardt, E. Characterization of Plasmodiophora brassicae populations from Alberta, Canada. Can. J. Plant Pathol. 28, 467–474 (2006).
Qu, C. et al. Genome-wide association mapping and Identification of candidate genes for fatty acid composition in Brassica napus L. using SNP markers. BMC Genom. 18, 232. https://doi.org/10.1186/s12864-017-3607-8 (2017).
Wang, N. et al. Genome-wide investigation of genetic changes during modern breeding of Brassica napus. Theor. Appl. Genet. 127, 1817–1829. https://doi.org/10.1007/s00122-014-2343-6 (2014).
Wu, Z. et al. Evaluation of linkage disequilibrium pattern and association study on seed oil content in Brassica napus using ddRAD sequencing. PLoS ONE 11, e0146383. https://doi.org/10.1371/journal.pone.0146383 (2016).
Malmberg, M. M., Shi, F., Spangenberg, G. C., Daetwyler, H. D. & Cogan, N. O. I. Diversity and genome analysis of Australian and global oilseed Brassica napus L. germplasm using transcriptomics and whole genome re-sequencing. Front. Plant Sci. 9, 508. https://doi.org/10.3389/fpls.2018.00508 (2018).
Glaubitz, J. C. et al. TASSEL-GBS: a high capacity genotyping by sequencing analysis pipeline. PLoS ONE 9, e90346. https://doi.org/10.1371/journal.pone.0090346 (2014).
Danecek, P. et al. The variant call format and VCFtools. Bioinformatics 27, 2156–2158. https://doi.org/10.1093/bioinformatics/btr330 (2011).
Browning, B. L., Zhou, Y. & Browning, S. R. A one-penny imputed genome from next generation reference panels. Am. J. Hum. Genet. 103, 338–348. https://doi.org/10.1016/j.ajhg.2018.07.015 (2018).
Cingolani, P. et al. A program for annotating and predicting the effects of single nucleotide polymorphisms, SnpEff: SNPs in the genome of Drosophila melanogaster strain w1118, iso-2, iso-3. Fly (Austin). 6, 80–92. https://doi.org/10.4161/fly.19695 (2012).
McLaren, W. et al. The ensembl variant effect predictor. Genome Biol. 17, 122. https://doi.org/10.1186/s13059-016-0974-4 (2012).
Bradbury, P. J. et al. TASSEL: software for association mapping of complex traits in diverse samples. Bioinformatics 23, 2633–2635. https://doi.org/10.1093/bioinformatics/btm308 (2007) (PMID:17586829).
Kimura, M. A simple method for estimating evolutionary rate of base substitutions through comparative studies of nucleotide sequences. J. Mol. Evol. 16, 111–120 (1980).
Raj, A., Stephens, M. & Pritchard, J. K. fastSTRUCTURE: Variational inference of population structure in large SNP data sets. Genetics 197, 573–589 (2014).
Tang, Y. et al. GAPIT Version 2: an enhanced integrated tool for genomic association and prediction. Plant Genome 9, 2–9. https://doi.org/10.3835/plantgenome2015.11.0120 (2018).
Aulchenko, Y. S., Ripke, S., Isaacs, I. & van Duijn, C. M. GenABEL: an R library for genome-wide association analysis. Bioinformatics 23, 1294–1296. https://doi.org/10.1093/bioinformatics/btm108 (2007).
Götz, S. et al. High-throughput functional annotation and data mining with the Blast2GO suite. Nuc. Acids Res. 36, 3420–3435. https://doi.org/10.1093/nar/gkn176 (2008).
Acknowledgements
The authors are grateful to Melissa Kehler, Victor Manolii, Md Mizanur Rahaman and the summer students Yasmina Bekkaoui and Kurtis Flavel for technical support. This work was funded by a competitive grant from SaskCanola and Agriculture and Agri-Food Canada under Canola Agronomic Research Program.
Author information
Authors and Affiliations
Contributions
F.Y. and A.D. conceived of and designed the study; A.D. and J.W. conducted the experiments; A.D., M.L., M.M.K. and Q.C. analyzed data; S.E.S., S.F.H., B.D.G. and G.P. provided important resources and facilities. A.D. drafted the manuscript. All authors reviewed the manuscript and approved the final draft.F.Y. and A.D. conceived of and designed the study; A.D. and J.W. conducted the experiments; A.D., M.L., M.M.K. and Q.C. analyzed data; S.E.S., S.F.H., B.D.G. and G.P. provided important resources and facilities. A.D. drafted the manuscript. All authors reviewed the manuscript and approved the final draft.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Dakouri, A., Lamara, M., Karim, M.M. et al. Identification of resistance loci against new pathotypes of Plasmodiophora brassicae in Brassica napus based on genome-wide association mapping. Sci Rep 11, 6599 (2021). https://doi.org/10.1038/s41598-021-85836-9
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-021-85836-9
- Springer Nature Limited