Introgression and monitoring of wild Helianthus praecox alien segments associated with Sclerotinia basal stalk rot resistance in sunflower using genotyping-by-sequencing

Sclerotinia basal stalk rot (BSR) and downy mildew are major diseases of sunflowers worldwide. Breeding for BSR resistance traditionally relies upon cultivated sunflower germplasm that has only partial resistance thus lacking an effective resistance against the pathogen. In this study, we report the transfer of BSR resistance from sunflower wild species, Helianthus praecox, into cultivated sunflower and molecular assessment of the introgressed segments potentially associated with BSR resistance using the genotyping-by-sequencing (GBS) approach. Eight highly BSR-resistant H. praecox introgression lines (ILs), H.pra 1 to H.pra 8, were developed. The mean BSR disease incidence (DI) for H.pra 1 to H.pra 8 across environments for four years ranged from 1.2 to 11.1%, while DI of Cargill 270 (susceptible check), HA 89 (recurrent parent), HA 441 and Croplan 305 (resistant checks) was 36.1, 31.0, 19.5, and 11.6%, respectively. Molecular assessment using GBS detected the presence of H. praecox chromosome segments in chromosomes 1, 8, 10, 11, and 14 of the ILs. Both shared and unique polymorphic SNP loci were detected throughout the entire genomes of the ILs, suggesting the successful transfer of common and novel introgression regions that are potentially associated with BSR resistance. Downy mildew (DM) disease screening and molecular tests revealed that a DM resistance gene, Pl17, derived from one of the inbred parent HA 458 was present in four ILs. Introgression germplasms possessing resistance to both Sclerotinia BSR and DM will extend the useful diversity of the primary gene pool in the fight against two destructive sunflower diseases.


Introduction
Cultivated sunflower (Helianthus annuus L.) is an important oilseed and confection crop worldwide. Fungal diseases caused by Sclerotinia sclerotiorum are of concern in sunflower production in the United States, as well as other parts of the world causing millions of dollars of crop losses annually [1]. S. sclerotiorum causes infection of sunflower plants at any growth stage and results in so-called Sclerotinia wilt or basal stalk rot (BSR), mid-stalk rot (MSR), and head rot (HR) diseases [2]. Sclerotinia BSR and HR are the most common sunflower diseases in the humid temperate production areas of Argentina, Europe, China, and the Northern Great Plains, where most of the U.S. sunflower crop is grown. The mode of infection for the two diseases differs. BSR is initiated by root infection from mycelia (unique to sunflower), while HR is caused by germination of airborne ascospores on sunflower capitula. Despite the common causal agent, the inheritance of resistance to Sclerotinia BSR and HR in sunflower is different based on the lack of a relationship between the two forms of the diseases [3,4]. Therefore, specialized screening nurseries and inoculation procedures are required for breeding of the two forms of Sclerotinia resistance, which effectively doubles the effort to combat the loss caused by the fungus. BSR is the predominant Sclerotinia disease of sunflower in the Northern Great Plains [3]. The sclerotia produced by the fungus plays a major role in the BSR disease cycle in the field. Under favorable environmental conditions, the mycelia from germinating sclerotia incite sunflower root infection. Typical BSR symptoms are revealed as a light-brown lesion girdling the stalk at soil level and extend up a few inches with the occasional presence of white, cottony mycelial growth visible under favorable moisture conditions. Over time, the fungus grows internally, destroying the pith and causing the plant to wilt and gradually die [2]. Chemical control of BSR is not feasible because the infection begins below ground. There is no effective fungicide registered in the U.S. and other countries for controlling BSR in sunflower. Therefore, breeding of resistant hybrids is the most efficient, economical, and environmentally friendly disease management approach.
Breeding for BSR resistance is challenging, since no immune germplasm has thus far been identified in sunflower or its close relatives. Nevertheless, various studies have demonstrated that resistance performance of diverse sunflower germplasms differs considerably [4,5,6] and the resistance is conditioned by multiple genes, each having a small effect [7,8,9]. Davar et al. [7] identified seven QTL associated with BSR resistance on LGs 1, 2, 4, 6, 8, 14 and 17 in a sunflower recombinant inbred line (RIL) mapping population of 116 individuals derived from the cross of PAC2/RHA266. Amouzadeh et al. [8] screened 99 RILs of the same population, but with a different fungal isolate and identified five QTL on LGs 1, 3, 8, 10 and 17. The effects of QTL for both studies were small (0.5-8%) with the QTL intervals ranging from 1.8 to 18.6 cM. Talukder et al. [9] identified two BSR resistance QTL on LG10 and LG17 using integrated data from multiple environments explaining 32 and 15%, respectively of the observed phenotypic variance. An additional four environment specific QTL were also identified on LGs 4,9,11 and 16, each explaining between 6 and 10% of the phenotypic variances. An association mapping effort identified two candidate genes on LG14 of the sunflower genome associated with BSR resistance [10]. Sclerotinia resistance in sunflower has traditionally been accomplished using classical genetic research and breeding efforts, often utilizing the available genetic variation in the primary gene-pool. However, the genetic base of the cultivated sunflower is narrow. Co-ancestry analysis has revealed that public sunflower inbred lines have originated from a small number of ancestral germplasm sources [11]. The consequence of the so-called founder effect, as defined by Ladizinsky [12], makes sunflower vulnerable to many biotic and abiotic stresses. Resistance present in the currently available cultivated sunflower gene-pool is not sufficient against the threat posed by the Sclerotinia [4,9,13]. Therefore, there is a growing need to improve Sclerotinia resistance in cultivated sunflower by diversifying its genetic variability utilizing the sunflower crop wild relatives. An abundance of sunflower crop wild relatives occupying a variety of habitats in the continental USA where they coevolved in the center of origin, are a valuable resource in the fight against biotic and abiotic stresses [14].
Cultivated sunflower belongs to the genus Helianthus, a member of the Asteraceae family consisting of 53 species, including 14 annual and 39 perennial [14]. All annual wild Helianthus species are diploid (2n = 2x = 34) and readily crossable with cultivated sunflower (except H. agrestis) with limited incompatibility, and homoeologous recombination occurs with relative ease. Genetic resistance has been identified in wild Helianthus species for sunflower rust (caused by Puccinia helianthi Schwein.), and downy mildew (caused by Plasmopara halstedii (Farl.) Berl. Et de Toni) and are routinely being deployed into cultivated sunflower as race-specific single dominant genes [15][16][17][18][19][20]. Earlier studies have repeatedly demonstrated high level of Sclerotinia resistance in the wild Helianthus gene-pool (reviewed by Seiler et al. [14]). Despite the devastating impact on the sunflower, it is apparent that wild Helianthus resources have not been adequately utilized for Sclerotinia resistance breeding. This limitation was partly due to the complex quantitative nature of the BSR resistance and the unavailability of efficient genomic tools to simultaneously assess multiple introgression regions in the cultivated sunflower background. However, the recent release of the sunflower reference genome sequence offers new opportunities for sunflower improvement by identifying genes of agronomic interest [21]. The use of high-throughput next-generation sequence (NGS) based genotyping-by-sequencing (GBS) technology in hybridization and introgression studies has increased the potential to identify single nucleotide polymorphism (SNP) variation in specific DNA targets across the entire genome for dissecting complex quantitative traits [22].
H. praecox Engelm. & A. Gray is an annual wild sunflower species, also known by the common name Texas sunflower. H. praecox has three subspecies: H. praecox subsp. praecox, H. praecox subsp. runyonii and H. praecox subsp. hirtus [23,24]. All three subspecies are endemic to the state of Texas in the USA, and grow on sandy soils of the coastal prairies. H. praecox and its hybrid progenies showed a high level of Sclerotinia resistance in various studies [25][26][27][28][29][30][31], making the species a valuable source for Sclerotinia resistance genes for introgressing into a cultivated sunflower background.
In the present study, we report the transfer of Sclerotinia BSR resistance from H. praecox into cultivated sunflower, as well as monitoring alien segments in the highly BSR resistant introgression lines (ILs) using GBS-derived SNP markers. Additionally, we report the integration of a broad-spectrum downy mildew (DM) resistance gene, Pl 17 , into BSR resistant ILs derived from one of the parents, HA 458. The germplasms developed and information generated in this study will help breeders expedite resistance breeding against two important sunflower diseases.

Plant materials
Five accessions of H. praecox (PI 413176, PI 435849, PI 468853, PI 435855, and PI 468847) were selected as BSR resistant donor parents identified by Block et al. [27,28]. These accessions were all collected from Texas, USA. Among the accessions, PI 413176 is subsp. praecox, PI 435849 and PI 468853 are subsp. runyonii, and PI 435855 and PI 468847 are subsp. hirtus. Three inbreed lines HA 89 (PI 599773), nuclear male sterile (NMS) HA 89 (PI 559477), and HA 458 (PI 655009) were used as cultivated sunflower sources. All these lines possess good agronomic traits, but they are susceptible to BSR disease. HA 89 was released in 1971 as an oilseed maintainer line by USDA-ARS and the Texas Agricultural Experiment Station. NMS HA 89 is a mutant developed by streptomycin treatment of HA 89 possessing a recessive gene, ms9 that controls male sterility [32]. It was released as nuclear male-sterile genetic stock in 1990 [33]. HA 458 was released in 2010 as a high oleic maintainer line carrying the DM resistant Pl 17 gene [34,35]. Two commercial sunflower hybrids, Croplan 305 and Cargill 270, were used as resistant and susceptible checks, respectively. Additionally, inbred line HA 441 was also used as a resistant control in each BSR screening test.

Crossing, backcrossing and generation advance
The selected five wild H. praecox accessions and NMS HA 89 were grown in the greenhouse. The first round of crosses were made in 2009 with H. praecox accessions as the male parent and NMS HA 89 as the female parent. A total of 2,131, 1,602, 1,679, 1,383 and 1,721 florets of NMS HA 89 were separately pollinated with pollen from H. praecox accessions, PI 413176, PI 435849, PI 468853, PI 468847 and PI 435855, respectively, to obtain F 1 seeds.
Basal stalk rot resistant F 1 plants were crossed with HA 458. The progenies from these crosses were termed BC 1 s. HA 89 was used as the recurrent female parent to backcross to the selected resistant BC 1 s. The BC 2 F 1 progenies were selfed and advanced to the BC 2 F 2 generation, followed by repeated selfing for four generations. The F 1 through BC 2 F 2 generations were screened for BSR resistance in the greenhouse, and only resistant progenies were advanced to the next generation. The BC 2 F 3 families and progenies of the following generations were evaluated for BSR resistance in the field nurseries during 2012 to 2015 with resistant progenies advanced to the next generation.

BSR screening in the greenhouse
The S. sclerotiorum fungal isolate NEB-274 was used for inoculum production of all greenhouse and field screening trials, as described by Qi et al. [36]. The seeds of each generation (F 1 to BC 2 F 2 ) along with the recurrent parent HA 89, and checks Cargill 270, HA 441, and the Croplan 305, were grown in the greenhouse in plastic flats each containing six rows of four 5.7 × 7.6 cm wells filled with Sunshine SB 100B potting compost (SunGro Horticulture, Bellevue, WA). The inoculation trays (54.6 × 34.3 × 10.2 cm) were prepared by spreading 120 g of inoculums on a layer of vermiculite placed on top of a fiberglass screen at the bottom of each tray. The inoculation trays were then placed in a dark and humid phytotron at~22˚C for three days before they were moved to the greenhouse. Three-week-old sunflower seedlings were carefully uprooted from the plastic flats and placed directly on the inoculums bed of the inoculation trays. The gaps at the base of the seedlings were filled with vermiculite to hold sufficient moisture when watered. The trays were incubated in the greenhouse at a soil temperature of 22-24˚C. The inoculated seedlings were visually inspected daily for disease symptoms and were scored at 14-18 days after inoculation (Fig 1). Sclerotinia BSR disease incidence (DI) is expressed as the percentage of dead and/or wilted plants.

Field experiments
The progenies of BC 2 F 3 through BC 2 F 5 families were grown and tested for BSR resistance in the field at Carrington, ND In all field trials, the hybrid sunflower Cargill 270 and Croplan 305 were used as the susceptible and resistant checks, respectively. Additionally, an USDA-ARS released inbred line HA 441 was also used as a resistant check. The seeds of the progeny lines, the recurrent parent, and the checks, were planted in 6-m long single row plots with 75-cm row spacing. In each plot, 25 seeds were sown per row with 20 plants kept after emergence for BSR evaluation. The field trials were laid out with a randomized complete block design with two replications in 2012 and 2013 and three replications in 2014 and 2015 per year and location for each ILs. Each field trial was artificially inoculated following the method developed by Gulya et al. [37]. Approximately 90 g of S. sclerotinia inoculum were applied for each entry in row-side furrows 5-6 weeks after planting at approximately the V6 growth stage [38]. Disease incidence (DI) was used as an index of BSR susceptibility measured at physiological maturity of the sunflower plants in the field at the R9 growth stage [38]. DI was calculated as the percentage of plants showing BSR symptoms in each row.

Statistical analysis
Analysis of variance (ANOVA) was performed for the replicated field tests data using the PROC Mixed of SAS version 9.4 [39]. Each location of individual year was considered as an environment. The genotypes were considered as fixed, while the remaining factors were treated as random effects using the model: where Y is the DI of the k th genotype tested in the j th replication of the i th environment, μ is the overall mean, l is the effect of the i th environment, b(l) is the effect of the j th replication nested in the i th environment, g is the genetic effect of the k th genotype, gl is the interaction effect of the k th genotype and i th environment, and e is the random experimental error. The comparison of the DI means among the different ILs was performed using the least significant difference (LSD) test [40] at the 5% level of significance.

Genotyping for tracking the introgressed alien chromsomal segments
To track the presence of H. praecox chromosomal segments, genotyping was performed using the GBS method described by Elshire et al. [41] for the selected ILs and the parental lines, HA 89 and HA 458. All five highly heterozygous H. praecox accessions were excluded from the GBS experiment. Leaves were collected from four greenhouse-grown young plants of each selected sunflower lines, bulked, and freeze-dried. Genomic DNA was isolated from the freeze-dried tissues using the 'DNeasy 96 plant kit' (Qiagen, Valencia, CA, USA). DNA concentrations were measured using a NanoDrop 2000 Spectrophotometer (Thermo Fisher Scientific, Wilmington, DE, USA). DNA samples of each line (~2.0 μg) were sent to the Institute of Biotechnology, Cornell University, Ithaca, NY for GBS analysis. In brief, genomic DNA samples of individual ILs and recurrent parents were digested separately with EcoT22I, a restriction endonuclease that recognizes a six base-pair sequence (ATGCAT). The digested DNA fragments were then ligated to two types of adapters: a barcoded adapter to identify each sample and a common adapter with an EcoT22I compatible sticky end. The DNA samples were pooled and PCR was performed to amplify the ligated products using primers complementary to the 'adapters' sequences. The PCR products were then purified and loaded for sequencing on an Illumina Hiseq 2000 (Illumina, USA). Sequencing produced an average of 1,785,943 good barcode reads for the two recurrent parents and an average of 2,187,155 good barcode reads for the eight ILs. SNPs were extracted using the GBS discovery pipeline version 3.0.166 implemented in TASSEL software [42]. Briefly, tag counts were generated from Illumina sequencing fastq files using the 'FastqTo-TagCountPlugin'. Tag counts were merged with 'MergeMultipleTagCountPlugin' (options: −c 3) and were aligned to the sunflower reference genome HA412.v1.0. (http://sunflowergenome.org) using the Burrow-Wheelers Alignment tool version 0.7.8-r455 [43] and converted into a 'TagsOnPhysicalMap' file for SNP calling using the TASSEL-GBS quantitative SNP caller. The GBS protocol identified 22,061 SNPs among the recurrent parents and the eight H. praecox ILs (S1 Table). The SNPs assigned to one of the 17 sunflower chromosomes were named with a prefix of S1 to S17, which corresponds to the respective chromosomes, followed by a number representing the physical position of the SNP on the genome. The SNPs that were unassigned to any of the 17 sunflower chromosomes, or had missing data in either of the parents, or showed polymorphism between HA 89 and HA 458 were removed, leaving a total of 10,530 SNP markers for further analysis.

Phenotype and genotype tests for DM resistance
Phenotypic screening of the DM resistance was performed in the parents, HA 89 and HA 458, and in the selected H.pra 1 to H.pra 8 of H. praecox ILs using the North America (NA) Plasmopara halstedii race 734. This is a highly virulent race identified in USA in 2010 [44]. HA 458 is a known carrier of DM R-gene, Pl 17 . Resistance for DM in these lines was tested using the whole seedling immersion method in the greenhouse under control conditions [35,45]. The susceptible plants produced numerous white fungal spores on the abaxial surface of the cotyledons and true leaves, while the resistant plants lacked spores. Genotyping of the parental lines, HA 89 and HA 458, and the eight selected ILs, H.pra 1 to H.pra 8 was performed using a simple sequence repeat (SSR) marker ORS963, and two single nucleotide polymorphism (SNP) markers, SFW04052 and SFW08268. These markers are tightly linked to the DM resistance gene Pl 17 [35]. A polymerase chain reaction (PCR) for the SSR and SNP markers was performed as described by Qi et al. [46] and Qi et al. [35], respectively. The PCR reactions were run on a Peltier thermocycler (Bio-Rad Lab, Hercules, CA, USA) and the products were size segregated in an IR 2 4300/4200 DNA Analyzer with denaturing polyacrylamide gel electrophoresis (LI-COR, Lincoln, NE, USA).

Hybridization and early generation selection for BSR resistance in the greenhouse
The F 1 seed set varied among the five H. praecox accessions used in this study ( Table 1). The highest number of F 1 seeds was produced in the crosses with the two accessions of subspecies hirtus (13.8% each), followed by the two accessions of subspecies runyonii (3.9 and 6.6%), while the lowest number of seed set was observed in subspecies praecox (1.2%).
Twenty-two to thirty-six F 1 seeds derived from the crosses with wild H. praecox accessions were grown and tested for BSR resistance in the greenhouse (

Evaluation of BC 2 F 2 populations for BSR resistance in the greenhouse
A total of eight BC 2 F 2 populations derived from the crosses with H. praecox subsp. runyonii accession PI 468853 with enough seed set were evaluated for resistance to BSR during the winter of 2011 and early spring of 2012. Either 48 or 72 plants in each population were tested for BSR resistance with a total of 480 BC 2 F 2 plants. Wide variation of DI was observed among the BC 2 F 2 populations, ranging from 10.4 to 69.4%, with a mean DI of 45.7% across eight BC 2 F 2 populations ( Table 3).
Sclerotinia BSR prevalence varied greatly across the years and/or locations. The Crookston 2013 environment was the most conducive for BSR in sunflower with a mean DI of 25.0%, which was followed by Grandin 2014 and Carrington 2012 environments with mean DI scores of 19.0 and 17.5%, respectively. Overall, 2015 was less conducive for BSR disease development as manifested by the low mean DI scores of 6.1 and 8.1%, respectively, at the Carrington and Grandin locations ( Table 4). The mean BSR DI of the eight H. praecox ILs ranged from 1.2 to Table 2

Tracking H. praecox alien segments in the ILs
Wide variation in the SNP distribution was observed throughout the sunflower genome of the ILs with the lowest in chromosome 6 (236 SNPs) and highest in chromosome 10 (1,034 SNPs) ( Table 5). Out of 10,530 filtered SNPs, 806 were polymorphic between the recurrent parents and one or more of the ILs (S4 Table) (Table 5). Although, the number of polymorphic SNPs varied across the genomes of the ILs, a few common introgression regions were detected (Fig 3). Overall, the introduced H. praecox segments in the eight ILs were mainly recovered on chro-   Table). Out of the 93 polymorphic SNPs on chromosome 10, 50 SNPs were shared among H.pra 2, H.pra 4, H.pra 5 and H.pra 7 (59.5% of the polymorphic SNPs), a common introgression  Fig 3). A total of 97 SNPs were shared between the two ILs, H.pra 3 and H.pra 8, accounting for 91.5% of the total resistant donor alleles recovered on chromosome 14 in these ILs (Table 6, S1 Fig). Additional introgression regions were also observed in some of the H. praecox ILs on chromosome 11 (Table 6, Fig 3).

DM resistance in the ILs
The sunflower inbred line, HA 458, used in the crossing scheme is resistant to DM disease conferred by the Pl 17 gene, effective against all virulent P. halstedii races currently identified in the USA [35,47]. The eight ILs were genotyped using the three DNA markers, SFW04052, ORS963, and SFW08268 that are linked to the Pl 17 gene. Pl 17 was mapped to a 2.9-cM interval Sclerotinia BSR resistance from wild species between SFW04052 and ORS963 [35]. SFW04052 was distal to Pl 17 at 2.1 cM, while ORS963 was proximal to Pl 17 at 0.8 cM in the genetic map. SFW08268 was downstream of ORS963 at 1.0 cM. Six of the eight ILs produced the same PCR pattern at three marker loci, while the remaining two lines, H.pra 4 and H.pra 6, had recombination events detected between SFW04052 and ORS963 (Table 7).  Table 5. Tracking of the alien segments introduced from H. praecox in the highly basal stalk rot resistant germplasm lines using single nucleotide polymorphism markers developed using the genotyping-by-sequencing approach.

LG2 (533)
LG3 LG4 LG5 (921) LG6 (236) LG7 (324) LG8 (697) LG9 (794) LG10 (1034) LG11 (558) LG12 (608) LG13 (652) LG14 (675) LG15 (474) LG16 (445) LG17 Phenotypic evaluation of the ILs was conducted using isolate of the NA P. halstedii race 734, and the disease responses of the ILs were generally consistent with the marker data, except for H.pra 4 ( Table 7). Lines H.pra 2 and H.pra 5 with all three DNA marker loci from HA 458 were homozygous resistant, and lines H.pra 1, H.pra 3, H.pra 7 and H.pra 8 with the three DNA marker loci from HA 89 were homozygous susceptible (Table 7). H.pra 6 was resistant to the disease although it was heterozygous at the SFW04052 marker locus. This could be the result of the HA 458 allele at ORS963 marker locus, which is the closest marker linked to Pl 17 at a genetic distance of 0.8 cM. The phenotype of H.pra 4 was resistant, although it had the allele from HA 89 at the ORS963 locus. This result suggests the possibility that a crossover event occurred between the Pl 17 gene and ORS963 marker.

Discussion
In the present study, we used five highly BSR-resistant H. praecox accessions, one from subspecies praecox and two each from subspecies runyonii and hirtus to transfer BSR resistance into the cultivated sunflower. As predicted, the F 1 hybrid seed set was very low for each cross (Table 1). In earlier studies, fewer than expected seed sets were reported in F 1 interspecies hybrids between cultivated sunflower and the wild H. praecox subspecies due to the meiotic chromosomal aberrations [48,49]. Although our crossing program began with five highly BSRresistant H. praecox accessions from three subspecies, we ended up with segregating progenies only from the cross involving PI 468853 H. praecox subsp. runyonii. Infertility of the segregating generations or reduced recombination between the chromosomes of the wild H. praecox subsp. praecox and H. praecox subsp. hirtus and cultivated sunflower might have eliminated progenies for BSR resistance evaluation.  BSR resistances have been successfully transferred from wild Helianthus species into cultivated sunflower background, and eight ILs have been developed from crosses of HA 89 with H. praecox through seven disease-screening cycles (F 1 to BC 2 F 5 ). A high disease pressure was used in the greenhouse screening trials in the early segregating generations (F 1 to BC 2 F 2 ) and only selected the highly resistant segregates to advance to the next generation. An intense selection pressure enhances the probability to recover the trait and favors the desired introgression fragment to be stable until the region becomes homozygous [50]. The ILs developed in this study largely showed stable BSR resistance across multi-location field screening trials in four years ( Table 4). The mean DI in the eight lines was significantly lower than those of the susceptible check, Cargill 270, and the recurrent parent, HA 89. Most of the ILs were either significantly more resistant than one or both of the resistant checks, except the IL, H.pra3, which had a DI similar to the checks ( Table 4). The prevalence of BSR disease varied across the field screening environments with the highest in 2013 and lowest in 2015, which became more evident from the BSR DI scores of the recurrent parent and both resistant and susceptible checks (Table 4). By contrast, with a few exceptions, the ILs consistently showed stable and superior resistance across environments, suggesting the successful transfer of novel Sclerotinia BSR resistance from wild H. praecox species. Nevertheless, variable level of BSR resistance has been observed among the eight ILs. BSR resistance in sunflower is controlled by quantitative genes with additive effects. When BSR resistance was transferred from wild species, the selected ILs might have integrated different partial resistance genes from wild species and resulted in different levels of resistance among selected ILs.
SNP variations are ubiquitous in the genome and are extremely suitable for a wide range of genomic studies [51]. GBS is an application of NGS technology that facilitates simultaneous discovery and genotyping of many SNP markers in crop genomes [41]. It is now routinely used for dissecting complex quantitative traits (for review Talukder et al. [9]; He et al. [52]) and, more recently, it has been demonstrated as a highly efficient tool for high-throughput tracking of introgressions [36,[53][54][55][56][57]. In our study, the GBS analysis discovered a total of 10,530 filtered SNPs of which 806 unique SNPs were polymorphic between recurrent parents and one or more of the ILs. Because the selection of ILs was performed under intense BSR disease pressure, the retained alien segments in the cultivated sunflower background are likely associated with Sclerotinia resistance. The polymorphic SNPs were distributed across the entire genome of the ILs, which was expected for a polygenically controlled quantitative trait.  (Table 6, S5 Table). While there were common introgressions detected among a few lines by shared polymorphic SNPs (  Table), suggesting transfer of novel Sclerotinia BSR resistance from wild H. praecox species. Sclerotinia BSR resistance QTL have been previously mapped using candidate gene association mapping [10] and in biparental mapping populations derived from cultivated sunflower background [7][8][9]. Talukder et al. [10] reported a strong association of Sclerotinia BSR with orthologs of the Arabidopsis thaliana COI1 (Coronatine Insensitive 1) gene, HaCOI1-1 and HaCOI1-2 located approximately at the positions 221.85 and 90.43 Mb regions, respectively, on the chromosome 14 of sunflower physical map, explaining 7.4% of phenotypic variation in the association mapping population. In our study, H. praecox alien segments were detected within~2 kb of the HaCOI1-1 gene in H.pra 3, H.pra 5 and H.pra 6, while alien segments were detected within~2 kb near the HaCOI1-2 gene in H.pra 3 and H.pra 8 ILs.
Talukder et al. [9] used GBS-derived SNP markers to map BSR resistance QTL in a sunflower recombinant inbred line (RIL) population developed from the cross of inbred lines HA 441/RHA 439. Two major QTL were identified on chromosomes 10 and 17 in multiple environments and each explained 31.6 and 20.2%, respectively, of the observed phenotypic variance in the RIL population. Our current study detected the H. praecox alien segment in H.pra 4, H.pra 5 and H.pra 7 within the tightly flanking SNP markers S10_281294015 and S10_288646223 (~7.35 Mb) of the BSR resistance QTL, Qbsr-10.1 on chromosome 10. Overall, a significant number of wild H. praecox alien segments was detected along the entire genome of the selected ILs each possessing higher levels of Sclerotinia BSR resistance. Some of these introgressions were detected in regions of previously identified BSR resistance QTL; the majority of which were unique and might be associated with new BSR resistance. A detail QTL study will elucidate the role of these alien segments in the underlying genetic mechanism of BSR resistance in these lines. Efforts are underway to evaluate the mapping population developed from wild H. praecox species for BSR resistance in locations across North Dakota and Minnesota.
One of the cultivated sunflower parents used in the current study, HA 458, is resistant to downy mildew, another major sunflower disease of global importance. HA 458 possesses a DM resistant gene Pl 17 that is highly effective against all known P. halstedii races thus far identified in the USA [34,35,47]. Although no additional effort was made to select DM resistance during IL development, H.pra 2, H.pra 4, H.pra 5, and H.pra 6 showed complete resistance to the highly virulent DM race 734 ( Table 7). The selected ILs with dual resistance against two important sunflower diseases, Sclerotinia BSR and DM, represent a valuable genetic source for disease resistance breeding in sunflower.
Despite the high level of BSR resistance available in the wild Helianthus species, adequate utilization of this invaluable resource has been limited in sunflower breeding due to the linkage drag and different incompatibility barriers between cultivated and wild species. Gene introgression from secondary gene-pools coupled with high-throughput tracking of introgressions presented here will provide a unique opportunity to expand the genetic base of cultivated sunflower by exploiting genetic variability present in wild species, as well as ensuring a continuous supply of new sources of resistance feeding into breeding pipelines to maintain the sunflower as a viable major global oilseed crop.