Mapping Quantitative Trait Loci (QTL) for Resistance to Late Blight in Tomato

Late blight caused by Phytophthora infestans (Montagne, Bary) is a devastating disease of tomato worldwide. There are three known major genes, Ph-1, Ph-2, and Ph-3, conferring resistance to late blight. In addition to these three genes, it is also believed that there are additional factors or quantitative trait loci (QTL) conferring resistance to late blight. Precise molecular mapping of all those major genes and potential QTL is important in the development of suitable molecular markers and hence, marker-assisted selection (MAS). The objective of the present study was to map the genes and QTL associated with late blight resistance in a tomato population derived from intra-specific crosses. To achieve this objective, a population, derived from the crossings of NC 1CELBR × Fla. 7775, consisting of 250 individuals at F2 and F2-derived families, were evaluated in replicated trials. These were conducted at Mountain Horticultural Crops Reseach & Extension Center (MHCREC) at Mills River, NC, and Mountain Research Staion (MRS) at Waynesville, NC in 2011, 2014, and 2015. There were two major QTL associated with late blight resistance located on chromosomes 9 and 10 with likelihood of odd (LOD) scores of more than 42 and 6, explaining 67% and 14% of the total phenotypic variation, respectively. The major QTLs are probably caused by the Ph-2 and Ph-3 genes. Furthermore, there was a minor QTL on chromosomes 12, which has not been reported before. This minor QTL may be novel and may be worth investigating further. Source of resistance to Ph-2, Ph-3, and this minor QTL traces back to line L3707, or Richter’s Wild Tomato. The combination of major genes and minor QTL may provide a durable resistance to late blight in tomato.


Introduction
Late blight (LB), caused by the oomycete Phytophthora infestans, (Montagne, Bary) is one of the most potentially devastating diseases of tomato in areas with high humidity and cool temperatures and can cause 100% crop loss in unprotected tomato fields or greenhouses. Because of its devastating economic impact, this disease has been the subject of intensified pathological and genetic research since the occurrence of the Irish potato famine in the 1840s. Genetic resistance to LB in tomato has been of interest for many years, and three major resistance genes have been identified in the red-fruited tomato wild species S. pimpinellifolium, including Ph-1, Ph-2, and Ph-3, which have been mapped to tomato chromosomes 7, 10, and 9, respectively. Ph-1 is a single dominant gene providing resistance to race T-0, but was rapidly overcome by new races of the pathogen. Ph-1 was mapped to the distal end of chromosome 7 using morphological markers [1]. However, no molecular marker associated with this resistance gene has been reported. Other genes Ph-4, and Ph-5-1 and Ph-5-2 have also been reported originating from LA1033 [2] and PSLP153 [3], respectively. However, Ph-4 turned out to be a QTL and Ph-5 is yet to be characterized for its biological role towards late blight control. Currently, P. infestans race T-1 predominates, rendering the resistance conferred by Ph-1 ineffective.  Late blight resistance comes from NC 1CELBR, and Fla. 7775 is susceptible to the late blight (see Materials and Methods for details). A qualitative resistance is expected to follow the discrete distribution pattern, whereas quantitative resistance is expected to follow a continuous distribution pattern. Late blight distribution patterns in F2, F3 and F4 populations are given in Figure 1. Late blight was normally distributed in an F2, F3, F4 generations and for overall average late blight scores for the entire population ( Figure 1). blight pressure in 2015. The average of all experiments scored was 1.9, with a range from 0.3 to 4.0 (Table 1). Although we started with 250 lines in 2011, by the time the study was completed in 2015, there were only 175 lines. In several lines, it was not possible to collect seeds because of disease severity. There was a high level of disease pressure in some lines, whereas others were completely healthy. Late blight resistance comes from NC 1CELBR, and Fla. 7775 is susceptible to the late blight (see Materials and Methods for details). A qualitative resistance is expected to follow the discrete distribution pattern, whereas quantitative resistance is expected to follow a continuous distribution pattern. Late blight distribution patterns in F2, F3 and F4 populations are given in Figure 1. Late blight was normally distributed in an F2, F3, F4 generations and for overall average late blight scores for the entire population ( Figure 1).   Table 2).

Quantitative Trait Loci (QTL) Analysis
Two major genes, Ph-2 and Ph-3 conferring resistance to late blight, are contributed from NC 1CELBR [18]. Ph-2 traces back to the Richter's Wild Tomato, whereas Ph-3 comes from L3707. QTL associated with late blight resistance in tomato were detected from chromosome 6, 8, 9, 10, and 12. There was one QTL detected from each of chromosome of 6 and 8 with 2.5 and 2.8, LOD scores, respectively, explaining about 2% to 8% of the total phenotypic variability (R 2 -value) (data not shown). QTL on chromosome 6 was contributed from P2 (Fla. 7775), since additive effect was positive, indicating that the level of disease resistance was high in the progenies with the alleles from this parent. QTL detected from chromosome 8 in 2011 and MRS, Waynesville 2015 were also contributed from P2 (Fla. 7775).
QTL from chromosome 9 were found consistent in all five environments, where Ph-3, a major gene conferring resistance to the late blight in tomato, is also located. It was not only consistent, but also a major QTL detected across the environments, with a LOD score of more than 42 explaining 67% of the total phenotypic variation. This QTL is located on about 67 cM position of the chromosome 9 from the telomere where molecular markers CL016855-0847, solcap_snp_sl_69978, and solcap_snp_sl_63704 are located (Table 3; Figure 2). Apparently, this was a major QTL associated with late blight resistance in tomato. Additive effect of the QTL was found to have −1.96, indicating that an individual allele contributed from the resistance parent (P1; NC 1CELBR), and that much disease resistance would increase.
Similarly, there was a major QTL associated with late blight resistance located on chromosome 10, with a LOD score of 7.4, explaining almost 14% of the total phenotypic variation, where Ph-2 is also located. This QTL was found to be located on about 63 cM of the chromosome. Molecular markers around this QTL are CL017176-0241, solcap_snp_sl_8855, solcap_snp_sl_8835, and solcap_snp_sl_8807 (Table 3; Figure 3). The additive effect of this QTL was estimated to be −0.56, indicating that with an individual allele, 0.56 disease resistance level would increase. Additive resistance alleles were contributed from P1 (NC 1CELBR) for this QTL.        There were also QTLs detected from chromosome 12 with the LOD scores of 3 and 2.6, respectively. However, the level of phenotypic variation explained by an individual QTL was only about 6%, indicating that these were minor QTL (Table 3). This QTL was found to interact with the QTL from chromosome 9 significantly (p < 0.05, Table 4). This may be an undetermined additional hypostatic gene(s)/QTL coming from L3707 or Ritcher's Wild Tomato. Based on the location of the QTL, there were two QTL from chromosome 12 located at a distance of 0.01 and 67 cM with the LOD scores of 3.13 and 2.12, respectively (Table 3). Resistance alleles were contributed from P1 (NC 1CELBR). The magnitude of disease resistance was about −0.30, indicating that this QTL may reduce the disease severity by 0.30 when scored at the scale of 0 to 5. Table 4. Evaluation of epistatic effect of quantitative trait loci (QTL) associated with late blight resistance from chromosome 9 and 12 in tomato. Table presents the direct outputs including model components and parameter estimates from the R-software analysis [19]. Model formula Y~Q1 + Q2.

Discussion
The objective of the present study was to map the genes and QTL associated with late blight resistance in a tomato population derived from intra-specific crosses. As presented in the Results section, there were two major QTL associated with the late blight resistance located on chromosomes 9 and 10. These had LOD scores of more than 42 and 6, explaining 67% and 15% of the total phenotypic variation, respectively. This indicated that these are the major genes/QTL from these two chromosomes. A co-dominant gene Ph-2 from chromosome 10 has been reported conferring resistance to late blight [4]. Similarly, Ph-3 conferring resistance to LB located on chromosome 9 has been reported [8,20]. Further, Ph-3 was fine-mapped in an interval of 0.5 cM between two molecular markers Indel_3 and P55, at a distance of 74 kb region [9]. These two molecular markers are from the region of TG591 RFLP marker, which is about 55 cM from the telomere. In the present study, we found the location of the race non-specific (which was found to be US-23 genotype of P. infestans) LB resistance QTL from these two chromosomes. It is possible that the gene/QTL from chromosome 9 is Ph-3, which is about 12 cM upstream from the fine-mapped location.
Ph-2 locus conferring partial resistance to late blight was mapped to chromosome 10 in an F2 population derived from Solanum pimpinellifolium (West Virginia 700 (WVa700) × HI7996) using amplified fragment length polymorphism (AFLP) molecular markers [4]. It was found to be located on the distal part (60 to 70 cM) of the chromosome, in an interval of 8.4 cM on the long arm of chromosome 10 near molecular markers CP105 and TG233. QTL detected in the present study was also from the same region, in fact, it was very close to TG223.
A race-specific resistance gene Ph-3 provides resistance to a broader range of isolates of Phytophthora infestans. The genotypes with Ph-1 and Ph-2 genes were susceptible against multiple isolates of Phytophthora infestans, as characterized by Kim and Mutschler [21]. It was believed that there may be an additional minor genetic allele in addition to Ph-3 to confer complete resistance to LB [21,22]. Chen, et al. [23] mapped the QTL associated with Ph-3 on chromosome 9 and also mapped an additional QTL from chromosome 2 derived from L3708 (S. pimpinellifolium). In the present study, we found the major QTL from chromosome 9 and 10. In addition to that, we also found a minor QTL from chromosomes 6,8 and 12, which may be novel minor QTL. These are important when it was widely believed that there should be additional QTL derived from L3707 and potentially from Richter's Wild Tomato, which needs to be verified in the future studies. A significant interaction between QTL from chromosomes 9 and 12 indicated that these two QTL may be inter-dependent or QTL from chromosome 12 (minor QTL) may play role in transcription of QTL from chromosome 9, which was found to be the major QTL.
While Ph-2 and Ph-3 are single genes conferring resistance to the late blight in tomato, a series of studies have shown to have quantitative resistance involved in tomato originating from wild relatives [24][25][26][27]. The source of resistance traces back to LA716, LA1777, and LA2099 [24,28,29]. In order to make the resistance durable, Li, Liu, Bai, Finkers, Wang, Du, Yang, Xie, Visser and van Heusden [28] have suggested the pyramiding of resistance gene and/or QTL from multiple species. Shandil, et al. [30] demonstrated that the level of resistance can vary in potato even if the resistance gene (R gene) was verified by PCR depending upon the genetic background of the recipient. Their conclusion was to investigate the role of other genes for achieving satisfactory level of resistance in potato using R gene. Jo, et al. [31] discuss about gene stacking approach to achieve the durable resistance in potato. Durable resistance has also been shown by combining the qualitative and quantitative resistance in potato in yet another study [32]. They indicate that there may be nucleotide-binding site-leucine-rich repeat (NBS-LRR) mediated resistance in potato. The role of NBS-LRR has also been shown to cause root knot nematode resistance in pepper [33,34].
Major genes have been identified from chromosome 9 and 10, and while this is consistent with past findings, the uniqueness of the present study is that we found some additional QTL from chromosome 6, 8, and 12 with the same source of resistance as Ph-2 and Ph-3. The presence of undetermined additional hypostatic gene(s)/QTL in L3707 is necessary to provide full resistance (R. Gardner, personal communication), and the present study has unraveled that to some extent. The pedigree of the present population traces back to the L3707 (S. pimpinellifolium), which is also the donor of Ph-3 [18]. It should be noted that the same line may be the donor of both the major gene Ph-3, as well as minor QTL detected on chromosomes 6, 8, and 12 in the conferring LB resistance in tomato. As mentioned before, Kim and Mutschler [13,21] and Irzhansky and Cohen [35] have reported the presence of additional late blight resistance derived from L3708 and L3707, respectively, and the presence of epistatic gene interaction for late blight resistance. This additional L3707/L3708-derived resistance is yet to be mapped and characterized. Since both lines have also been reported to be the source of Ph-3 gene, it is worth characterizing.

Plant Materials
Tomato breeding line NC 1CELBR was developed at North Carolina State University (NCSU) by Dr. R. G. Gardner. It is a large-fruited fresh-market tomato breeding line with a determinate growth habit and is resistant to LB conferred by Ph-2 and Ph-3 genes [18]. Seeds of the susceptible line Fla. 7775 were kindly provided by Dr. Jay Scott, University of Florida. Despite other similar characterisrtics, contrasting LB reactions in NC 1CELBR and Fla. 7775 provided ideal materials to develop a population for genetic mapping studies. Crosses were made in the fall of 2009 at the Mountain Horticultural Crops Research and Extension Center, NCSU, Mills River, NC. The F 2:3 families were developed in the spring of 2010 by selfing the F 2 . Subsequently, the F 2:3 population was developed and used for single nucleotide polymorphism (SNP) marker analysis, QTL mapping, and phenotypic evaluation in the field.

Phenotyping for Disease Resistance in Field in 2011
To evaluate resistance to LB in the field, the experiment was conducted in 2011. Seeds were sown in 72 cell flats (56 × 28 cm 2 ) in the first week of May. Transplants, at about 6 weeks, were planted by hand in the field. Transplants were planted 45 cm apart in a planting row, with 150 cm between rows. The soil was a clay-loam and the natural day light photoperiod was about 14/10 h with 25-30 • C high and 14-16 • C low day/night temperatures. In the first week of June 2011, the 183, four-week-old greenhouse grown transplants of the F 2 and F 1 hybrid (NC 10175), along with susceptible checks (Fletcher, NC123S and NC 30P) and resistant checks (NC 2CELBR, NC 25P, Plum Regal and Mountain Merit), and parents (NC 1CELBR and Fla. 7775) were planted at the Mountain Research Station, Waynesville, NC. This field site was chosen because P. infestans inoculum occurs naturally each year. Parents and the F 1 were planted as a check to make sure that disease developed well in the susceptible parent, and that the resistant parent was healthy even under high inoculum pressure. In the field, a disease rating was performed on the scale of 0 to 5, where 0 = no disease symptoms on the leaf surface area, 1 = symptoms spread over about 20% of the leaf surface area, 2 = symptoms spread over 21-40% of the leaf surface area, 3 = symptoms spread over about 41-60% of the leaf surface area, 4 = symptoms spread over 61-80% of the leaf surface area, and 5 = symptoms spread over 100% of the leaf surface area. Plants that showed defoliation ≤40% were considered as LB resistant and plants exhibited defoliation ≥40% were considered as LB susceptible.

Phenotyping for Disease Resistance in the Field in 2014 and 2015
In 2014 and 2015, the F 2:3 and F 2:4 families, parents (NC 1CELBR and Fla. 7775) and control lines were sown into seeding trays in a standard seeding mix (2:2:1 v/v/v) peat moss:pine bark:vermiculite with macro-and micro-nutrients (Van Wingerden International Inc., Mills River, NC, USA). After 10 days, seedlings were transplanted to 72-cell flats (56 × 28 cm 2 ). Six plants per genotype were planted with two replications, and the same experiment was conducted in MHCREC, NC.
As the plants were infected with natural inoculum, plants were scored on a scale of 0 to 5, where 0 = no disease symptoms on the leaf surface area, 1 = symptoms spread over about 20% of the leaf surface area, 2 = symptoms spread over 21-40% of the leaf surface area 3 = symptoms spread over about 41-60% of the leaf surface area, 4 = symptoms spread over 61-80% of the leaf surface area, and 5 = symptoms spread over 100% of the leaf surface area. Plants that showed defoliation ≤40% were considered as LB resistant and plants exhibited defoliation ≥40% were considered as LB susceptible.

DNA Isolation and SNP Genotyping
Genomic DNA of young leaf tissues of each line and parent were extracted using DNeasy Plant Mini Kit (Qiagen Inc., Valencia, CA, Spain). A Nano Drop (Model ND-2000, Thermo Scientific Inc., Wilmington, DE, USA) was used to quantify each DNA sample. Approximately 50 ng/µL of DNA was prepared from each sample for SNP genotyping. We used an optimized subset of 384 SNPs markers that were derived from the 7,725 SNP array developed by the Solanaceae Coordinated Agricultural Project (SolCAP) [36,37]. The subset of markers was selected based on polymorphism rates among six fresh market tomato accessions including Fla.7776, Fla.8383, NC33EB-1, 091120-7, Fla. 7775, and NC 1CELBR. In addition, genetic position in the genome based on recombination [36] and physical position were considered as important selection criteria to insure genome coverage. These 384 SNPs were analyzed using the KASP genotyping platform (LGC Genomics, Beverly, MA, USA).

Phenotypic Data Analysis
The disease rating values were calculated from replicated trials to measure resistance to LB in 2011 from the 183 F 2 plants in the field trial and in 2014 and 2015 from the F 2:3 and F 2:4 populations in the field trials. Data from the field experiments were subjected to analysis of variance (ANOVA) using Proc Mixed in SAS 9.3 [38]. Correlation analysis was performed between different environments using PROC CORR procedure of SAS.

QTL Analysis
Out of the 384 SNP markers, only 184 were polymorphic between the two parental lines. NC 1CELBR and Fla. 7775, were used for QTL analysis and to develop genetic maps [39]. Recombination frequencies of the map were converted into genetic distance (cM) using the Kosambi mapping function and calculation of genetic distance between two adjacent SNP was performed [40].
QTL analysis for late blight was carried out by Composite Interval Mapping (CIM) by using QTL Cartographer v 2.5 software [41]. A default threshold of 2.5 was used to declare the presence of QTL in all environments. We used 5 cM scanning steps for the detection of QTL. The relative contribution of genetic component was calculated, and described as the proportion of the phenotypic variance explained. QTLs explaining more than 10% of the phenotypic variance were considered as major QTL, and QTL found in at least two environments were considered to be consistent. Data analysis was repeated by using R software to confirm the presence of QTL and their interaction [19].
Any QTL within 5 cM distance on the same chromosome were regarded as a single QTL. The potential biological function of the SNP markers that were associated with each QTL for resistance to LB were inferred using an in silico approach. The SNP marker sequences were blasted against coding regions of Arabidopsis thaliana, rice (Oryza sativa L.) and tomato (S. lycopersicum L.) databases, genes essential for metabolic pathways, and plant defense-related were identified.