QTL Mapping of Resistance to Pseudoperonospora Cubensis Clade 2, Mating Type A1, in Cucumis Melo and Dual-clade Marker Development

doi:10.21203/rs.3.rs-1735122/v1

Download PDF

Research Article

QTL Mapping of Resistance to Pseudoperonospora Cubensis Clade 2, Mating Type A1, in Cucumis Melo and Dual-clade Marker Development

https://doi.org/10.21203/rs.3.rs-1735122/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 03 Apr, 2023

Read the published version in Theoretical and Applied Genetics →

You are reading this latest preprint version

Pseudoperonospora cubensis, causal organism of cucurbit downy mildew (CDM), causes severe necrosis and defoliation on Cucumis melo (melon). A recombinant inbred line population (N = 169) was screened against an isolate of P. cubensis (Clade 2 / mating type A1) in replicated greenhouse and growth chamber tests. SNPs (n = 5,633 bins) identified in the RIL population were used for quantitative trait loci (QTL) mapping. A single major QTL on chromosome 10 was consistently associated with resistance across all tests, while a second major QTL on chromosome 8 was identified only in greenhouse tests. These two major QTL were identified on the same chromosomes (8 and 10) but in different locations as two major QTL previously identified for resistance to P. cubensis Clade 1 / mating type A2. Kompetitive allele specific PCR (KASP) markers were developed for these four major QTL and validated in the RIL population through QTL mapping. These markers will provide melon breeders a high-throughput genotyping toolkit for development of melon cultivars with broad tolerance to CDM.

Melon

cucurbit downy mildew

Pseudoperonospora cubensis

QTL

KASP

This is the first identification of QTL underlying resistance in Cucumis melo to an isolate of Pseudoperonospora cubensis identified as Clade 2 / mating type A1.

Cucurbit downy mildew (CDM) is caused by the obligate oomycete plant pathogen Pseudoperonospora cubensis (Berk. et Curt.) Rostov. P. cubensis causes severe chlorosis and necrosis in all six major commercially cultivated cucurbit species. In the eastern United States, windblown sporangia travel northward each year, lagging about a month behind cucurbit planting dates, and land on leaf surfaces to initiate infections (Ojiambo et al. 2015). Many cucurbits are planted twice in a single growing season, with CDM more severe in the summer and fall months than in the spring (Keinath 2019). Management of P. cubensis relies heavily on fungicide applications due to insufficient host resistance in C. melo. There are several instances of fungicide resistance developing in P. cubensis, increasing the need for developing resistant cultivars (Keinath et al. 2019; Olaya et al. 2009; Thomas et al. 2018).

A CDM epidemic in the eastern United States in 2004 was caused by the introduction and spread of a new P. cubensis population that was genetically similar to Clade 2 isolates originating from Eurasia (Holmes et al. 2015; Runge et al. 2011; Quesada-Ocampo et al. 2012). A recent survey of P. cubensis populations utilizing C. melo sentinel plots from nine states identified two genotypes causing infection: 88% of isolates collected were Clade 1 / mating type A2 and 12% were Clade 2 / mating type A1 (Toporek and Keinath 2021). The inverse association of clade and mating type appears to be 100% in recovered isolates from C. melo, but alternative combinations have been observed in isolates from other cucurbits in South Carolina (Toporek, unpublished).

The two P. cubensis clades, potentially cryptic species, are host adapted, and populations of certain clades harbor specific, heritable fungicide resistances (D’Arcangelo et al. 2020; Runge et al. 2011; Wallace et al. 2020). Thus, these clades should be considered separate populations of interest when screening cucurbit germplasm for resistance. A recent population genetic analysis with seven gSSRs further supports this speciation within P. cubensis (Nowicki et al. 2021). An analysis of RXLR effectors determined that isolates of these two P. cubensis clades had both shared and unique effectors (Purayannur et al. 2020). Oomycete mating type has also been associated with the breakdown of host resistance, as was the case with Phytophthora infestans, when the introduction of a new mating type from Mexico caused a resurgence of late blight on potato and tomato in the United States (Goodwin 1997). For this reason, P. cubensis mating type may be an important consideration when screening C. melo germplasm.

Inheritance of resistance to P. cubensis in C. melo has been reported as polygenic, single dominant, or incompletely dominant (Epinat and Pitrat 1989; Perchepied et al. 2005; Thomas et al. 1988;). Recently, nine QTL for resistance to a Clade 1 / mating type A2 P. cubensis isolate were identified in a recombinant inbred line (RIL) population, including two major QTL on chromosome 8 and 10 derived from the resistant parent, MR-1 (Toporek et al. 2021). A promising candidate gene was identified as an MLO gene as a loss-of-function mutation in these genes has been shown to confer resistance to powdery mildew in barley, Arabidopsis thaliana, and pepper (Acevedo-Garcia et al. 2014). MR-1 was derived from PI 124111 (Thomas 1986), a CDM-resistant accession introduced from India, where Clade 2 isolates predominate (Runge et al. 2011).

Kompetitive allele specific PCR (KASP) markers are effective molecular tools that plant breeders can use to rapidly introduce genomic regions of resistance into new plant genotypes. These markers are developed around single nucleotide polymorphisms (SNPs) within the QTL of interest so that plant material can be evaluated with high-throughput screening for the presence of resistant alleles, alleviating the need to phenotype in earlier breeding steps. KASP markers have been successfully developed across qSulf-1, a QTL for sulfur resistance identified from the same MR-1 ×AY mapping population (Branham et al. 2020). KASP markers also were developed recently in C. melo and Cucumis sativus for powdery mildew (Cao et al. 2021).

The objective of this study was to 1) phenotype the segregating RIL population for resistance to P. cubensis Clade 2 / mating type A1, 2) identify whether unique QTL underly resistance to Clade 2 / mating type A1, and 3) develop and evaluate KASP markers for major QTL associated with both clades/mating types of P. cubensis.

P. cubensis isolate identification and inoculum production. An isolate of P. cubensis, SS-01,was recovered from a symptomatic leaf of cucumber cultivar Silver Slicer in June 2019 at the Clemson University Coastal Research and Education Center, Charleston, SC. Inoculum of SS-01 was increased and maintained as in Toporek et al. (2021), with the exception that sporangia were spray-inoculated onto 2-week-old leaves of ‘Silver Slicer’ grown in 50-cell trays (Pro-trays, Hummert International, Earth City, MO) instead of ‘Waltham’ butternut squash. Five lesions from the initial leaf were randomly selected, and DNA was extracted using a Synergy™ Plant DNA Extraction Kit (OPS Diagnostics, Lebanon, NJ) following the manufacturer’s instructions.

DNA concentrations were quantified using a Nanodrop ND 1000 spectrophotometer and NanoDrop 2.4.7c software (NanoDrop Technologies Inc., Wilmington, DE) and normalized to 7.5 ng/μL for downstream applications. The qPCR protocol developed by Rahman et al. (2021) was used to identify to which clade an isolate belonged, with fluorescence quantified using a Stratagene Mx3005P (Agilent Technologies, Santa Clara, CA) quantitative PCR system. The conventional PCR protocol developed by Dr. P. S. Ojiambo (personal communication) was used to determine mating type. Mating type gene amplifications were performed with a PTC-200 thermal cycler (Bio-Rad, Hercules, CA), and amplicons were visualized on a 1.5% agarose gel to verify amplicon size. Amplicons were sequenced bi-directionally by Sanger sequencing using Big Dye V3.1 chemistry and run on ABI 3730XL instruments (Functional Biosciences, Madison, WI). Resulting sequences were trimmed and aligned to mating type A1 and A2 reference gene sequences in Geneious R11 (Geneious, Auckland, NZ).

Plant materials. The U.S. Vegetable Laboratory (U.S. Department of Agriculture, Agricultural Research Service) generated a cross between CDM resistant breeding line MR-1 and susceptible cultivar Ananas Yoqne’am (AY) (Paris et al. 2013) to create a RIL population (F₆–F₁₁) through single seed descent from an F₂ population, resulting in 169 lines that segregate for resistance to P. cubensis. (Daley et al. 2017, Branham et al. 2018). A hierarchical nested design was used to screen seedlings of each RIL against SS-01 in two separate experiments in a greenhouse and two separate experiments in a growth chamber (Toporek et al. 2021).

Seeds of each of the 169 RIL, MR-1, AY, and F₁ hybrids were sown into 50-cell propagation trays and placed on heating pads for 48 h at 32ºC to encourage uniform germination. Experiments included two replications of five plants of each RIL, with two replications of five plants of MR-1, AY, and the F₁ hybrid included as controls. Five days prior to inoculation, seedlings were fertilized with Peters 20-20-20 (N-P-K) water-soluble fertilizer (JR Peters Inc., Allentown, PA) prepared at 2 g/L to promote a dark green leaf to increase contrast between healthy and diseased tissue. Both growth chamber experiments were conducted at 25ºC and 65% relative humidity (RH) with a 12 h photoperiod under LED lights. Greenhouse experiment 1 was seeded October 19, 2020 and rated November 9, 2020. Temperatures ranged from 22°C to 38.5°C, with an average of 25.3°C, and RH ranged from 24.6% to 82.8%, with an average of 62.8%. Seeds for greenhouse experiment 2 were sewn November 23, 2020 and rated December 21, 2020. Temperatures ranged from 16°C to 35°C, with an average of 21.6°C, and RH ranged from 20.5% to 85.5%, with an average of 57.7%.

Phenotyping. Inoculations in all four experiments were performed using freshly produced SS-01 sporangia collected from diseased 2-week-old ‘Silver Slicer’ seedlings as previously described (Toporek et al. 2021). For each seedling, the first true leaf was inoculated with 2 ×10⁴ sporangia/ml and placed into the humid chamber for 24 h, then transferred to either the greenhouse or the growth chamber. Symptoms were allowed to develop for 7 days, then the first true leaves were carefully removed with scissors, placed on a dark blue felt background, and photographed from above with a mounted camera. Photographs of each replication of an inoculated seedling were loaded into the image analysis software program Assess 2.0 (American Phytopathological Society, St. Paul, MN), and the total percentage of discolored leaf area of each leaf was quantified.

Statistical analysis. Phenotypic means were collected across all seedlings of each RIL from each test, excluding any RIL with less than two seedlings per test.Datasets were created for each of the four independent tests, combined data from both tests in an environment, and all tests combined from both environments, resulting in seven different phenotypic datasets. These datasets were tested for normality using Shapiro-Wilk tests in JMPPro, Ver. 14 (SAS Institute Inc., Cary, NC). Datasets were non-normal and correlations among tests were analyzed using Spearman’s correlation coefficients in JMP Pro. To calculate variance components, a mixed model was fit using a restricted maximum likelihood algorithm in the lme4 package in R (Bates et al. 2015). Model variables were treated as random effects: RIL (varRIL), environment (greenhouse or growth chamber) (varE), test (varT), the interaction of RIL and test, the interaction of RIL and environment, the interaction of test and environment, the three-way interaction of RIL × environment × test, and block nested within test. Broad-sense heritability (H²) was calculated as the variance of RIL divided by the overall phenotypic variance (Branham et al 2019).

QTL mapping. In conjunction with the C. melo genetic map created by Branham et al. (2018) for this population, comprised of 5,663 binned SNPs spanning all 12 C. melo chromosomes, the seven phenotypic datasets were individually used to identify QTL for CDM resistance. Cross objects of the genetic map and phenotypic data were analyzed using the ‘qtl’ package in R (Broman et al. 2003; R core team 2021). Standard interval mapping with Haley–Knott regression was performed with the ‘scanone’ function (Haley and Knott 1992), followed by forward selection, to determine the optimal number of QTL for downstream modelling and to generate LOD profile figures. The function ‘stepwiseqtl’ was then used to identify the optimal model with five or less QTL for each trait using an automated forward and backward selection algorithm (Broman and Sen 2009; Manichaikul et al. 2009). Genome-wide LOD significance for each dataset was determined with the function ‘scantwo’ based on 1,000 permutations.

KASP Marker Development and Validation. Whole genome resequencing of the parents generated 304,864 SNPs (Branhma et al. 2020) which were used to design KASP markers for the major QTL associated with resistance to P. cubensis Clade 1 / mating type A2 (Toporek et al. 2021) and Clade 2 / mating type A1. These SNPs had been assigned functional roles according to the reference genome annotation (Garcia-Mas et al. 2012) using ANNOVAR version 2017 (Wang et al. 2010). Candidate genes could be identified within the 1.5 LOD were prioritized by the functional annotation of the SNPs.

Ten KASP markers per QTL were designed across qPcub-8.3 and qPcub-10.3-10.4, the major QTL identified in this study, as well as qPcub–8.2 and qPcub–10.1, the major CDM resistance QTL identified in Toporek et al. (2021) associated with resistance to P. cubensis Clade 1 / mating type A2. SNPs were chosen that flanked the QTL peak, were greater than or equal to 50 bp away from each other, preferentially not part of an intron, and within different haplotype blocks, if possible. Preference was also given to SNPs that were non-synonymous or fell within a promotor region.

These SNPs, with the base pair of both parental alleles and their flanking 60 bp regions defined, were supplied to LGC Genomics (Teddington, Middlesex, UK) for “KASP™ by design” services (Online Resource 1). LGC Genomics then developed KASP marker assays for each SNP. PCR reactions (5 µl volume) consisted of 0.07 µl of primer mix (LGC Genomics; fluorophore-labeled allele-specific forward primers and a reverse primer), 2.5 µl of 2X master mix (LGC Genomics) and 20 ng of sample DNA. A PTC-200 thermal cycler was used for a touchdown PCR reaction with a 94°C hot-start activation step for 15 min, then ten cycles of 94°C (20 s) and a starting annealing temperature of 61°C, dropping by 0.6°C each cycle. Twenty-six additional cycles of 94°C for 20 s and 55°C for 60 s followed the touchdown steps. Fluorescence was visualized on a Stratagene Mx3005P and genotypes were called using MxPro v4.10 software.

KASP markers were initially screened using genomic DNA extracted from 2-week-old seedlings of MR-1, AY, and the F₁ hybrid using a Synergy 2.0 Plant DNA Extraction Kit, following the manufacturer’s instructions. DNA was quantified using a Nanodrop ND 1000 spectrophotometer and NanoDrop 2.4.7c software and normalized to 20 ng/ul. Each KASP marker was evaluated against six replicates of each parent and the F₁ hybrid, as well as six non-template controls, using the previously described PCR protocol. Markers were evaluated based on genotypic clustering profiles. Successful markers were then screened against the entire 169-RIL population and resulting genotypes were added to the GBS-generated genetic map. QTL mapping was conducted as previously described.

Downy mildew resistance. Pseudoporonospora cubensis isolate SS-01 was categorized as Clade 2 and mating type A1 through established molecular protocols. Disease response of the control plants of MR-1, AY, and F₁fell in the tails and peak of the population distribution, respectively, in all four tests (Fig. 4.1). Averaged across all tests, disease severity in the RIL population ranged from 2.87 to 38.63%, with a mean of 11.78%. Disease severity ranged from 2.27% to 50.11%, with a mean of 9.31% across the green house tests. Averaged across growth chamber tests, disease severity ranged from 2.25% to 70.05%, with a mean of 13.54%.

The distributions of phenotypic data (average percentage diseased leaf area) were non-normal across RILs (P = <0.0001) in all datasets (n=7). This non-normality warranted the use of the non-parametric Spearman’s correlation to examine correlations between groups of tests. Spearman’s correlation coefficients among tests and combinations of tests were significant in all comparisons, except when compared with growth chamber test 1. Both greenhouse tests were less correlated (r = 0.47, P = <0.0001) than greenhouse test 1 and growth chamber test 2 (r = 0.59, P = <0.0001), but more correlated than greenhouse test 2 and growth chamber test 2 (r = 0.40, P = <0.0001). When averaged together, data from both greenhouse tests and both growth chamber tests were significantly correlated (r = 0.46, P = <0.0001), but more correlated when growth chamber test 1 was discarded (r = 0.62, P = <0.0001), and not correlated when only compared to growth chamber test 1 (r = 0.06, P = 0.445).

Among variance components, the highest attributable variance was observed for the factors RIL (14.2%) and RIL × test × environment (14.8%), with a large proportion (57.7%) attributed to error. Environment and test accounted for 4.2% and 3.3%, respectively. H² of resistance to P. cubensis Clade 2 / mating type A1 was moderately low at 0.32.

QTL for cucurbit downy mildew resistance. RIL disease severity means were used as the phenotypes in QTL mapping. Genome-wide significance thresholds for means across tests, across greenhouse tests, across growth chamber tests, greenhouse test 1, greenhouse test 2, growth chamber test 1, growth chamber test 2 had LOD scores of 3.19, 3.01, 2.94, 2.89, 3.00, 3.12, and 3.13, respectively. Across all tests in both environments, the QTL model had a LOD score of 11.93. Greenhouse test 1 and means of both greenhouse tests 1 and 2 had higher LOD scores of 18.10 and 19.41, respectively. Growth chamber test 2 and means of both growth chamber tests 1 and 2 had the lowest LOD scores, 3.97 and 8.14, respectively.

Five QTL were identified with significant LOD scores across three chromosomes and named with the following rules as in Toporek et al. (2021): q(Pcub for Pseudoperonospora cubensis)-(chromosome).(unique identifier per chromosome) (Table 4.2; Fig. 4.2). All resistance alleles were non-epistatic and contributed by the resistant parent MR-1. qPcub-8.3 (LOD=9.1) and qPcub-10.5 (LOD=3.7) were the only two QTL identified using phenotype data means from all tests, explaining 22.5% and 8.3% of variation in the population (Table 4.2). qPcub-8.3 was the only QTL consistently observed in each test in which a QTL could be identified. qPcub-10.3 (LOD=9.4) and aPcub-10.4 (LOD=10.6), likely the same QTL, explained at most 21.1% of the variation observed in greenhouse tests only. qPcub-5.2 (LOD=3.7) explained 15.7% of variation observed across growth chamber tests only. No significant QTL were identified in greenhouse test 2 and growth chamber test 1.

Within the major QTL identified on chromosome 8, qPcub-8.3, 655 genes were identified over a 16.5 cM range between 8.6 cM and 25.1 cM (Table 4.2; Online resource 1). When narrowing results to just those genes within the range of observed peaks across different tests for qPcub–8.3, two genes were identified: MELO3C008820, a putative disease resistance gene nucleotide-binding site leucine-rich repeat (NBS-LRR) analog, and MELO3C008838, a receptor-like kinase (RLK). Within the major QTL identified on chromosome 10, qPcub–10.3 and qPcub–10.4, 7 and 16 genes were identified, respectively, within a much narrower 0.1 cM range. Within these genes, MELO3C011814 was identified as another putative RLK.

KASP Markers. KASP markers (n = 40) were designed using whole genome resequencing SNP data for parents MR-1 and AY, 10 per major QTL. MR-1, AY, and F₁were polymorphic and clustered accordingly with 28 of the developed markers, whereas 14 formed irregular genotype clusters or were monomorphic. This corresponded to 8, 4, 7, and 9 polymorphic markers that spanned qPcub-8.3, qPcub-10.3-10.4, qPcub-8.2, and qPcub-10.1, respectively (Table 4.3). QTL mapping was re-run using combined phenotype data from all tests of their respective studies for qPcub-8.3 and qPcub-10.1, combined growth chamber phenotype data for qPcub-8.2, and combined greenhouse data for qPcub-10.3-10.4. Mapping with the addition of these polymorphic KASP markers resulted in a narrowing of the physical distance of the LOD interval of each QTL.

For qPcub-8.3, the nearest marker CDM_C2_8_2115936 lay 0.168 Mb from the peak at 21.3 Mb. This peak was located within the same haplotype block as the marker. All other markers were located within the new 1.5 LOD interval. Adding KASP markers improved resolution of qPcub-8.3 by 0.2 Mb.

For qPcub-10.3-10.4, the peak was between 4.62 Mb and 4.65 MB. The nearest flanking marker was CDM_C2_10_4628519. Only CDM_C2_10_4651132 was also located within the new 1.5 LOD interval. Adding KASP markers improved resolution of qPcub-10.3-10.4 by 0.07 Mb.

For qPcub-8.2, the peak was between 27.42 Mb and 27.57 Mb. The nearest flanking marker was CDM_C1_8_27375196. Five markers were located within the new 1.5 LOD interval CDM_C1_8_26523285, CDM_C1_8_27362711, CDM_C1_8_27375196, CDM_C1_8_27595209, and CDM_C1_8_27815348. Adding KASP markers improved resolution of qPcub-8.2 by 0.2 Mb.

For qPcub-10.1, the new peak was located at 0.899 Mb, exactly at KASP marker CDM_C1_10_899379. The next closest marker, CDM_C1_10_899481, was 102 bp away. Six additional markers were located within the new 1.5 LOD interval CDM_C1_10_331322, CDM_C1_10_468391, CDM_C1_10_470297, CDM_C1_10_471835, CDM_C1_10_638624, and CDM_C1_10_720733. Adding KASP markers improved resolution of qPcub-10.1 by 0.3 Mb.

This study successfully identified five CDM resistance QTL, with one major QTL, qPcub–8.3, consistent across all experiments and another, qPcub–10.3-10.4, only identified in greenhouse experiments. Between environments, higher mean disease severity was observed in the growth chamber than in the greenhouse, which may partially explain why an additional QTL was identified in this environment. The opposite effect of environment was observed in Toporek et al. (2021), where an additional major QTL was observed only in growth chamber environments. The lower disease severity in the greenhouse also may be explained by greater variation in daily relative humidity and temperature, with daytime temperatures sometimes exceeding those favorable for CDM development, whereas the growth chamber temperature was kept constant within a favorable range (Cohen 1981). The environmental dependence of CDM host resistance has previously been reported for species of Cucumis (Kozik et al. 2013; Li et al. 2018; Toporek et al. 2021). Despite the consistency of environment, growth chamber test 1 was unexpectedly uncorrelated with growth chamber test 2, which may be indicative of microenvironments within the growth chamber.

Overall, disease severity from tests within and between environments were less correlated in this study using a Clade 2, mating type A1 isolate than previously described for a Clade 1, mating type A2 isolate, potentially indicating a greater sensitivity to environment for isolate SS-01 (Toporek et al. 2021). Nevertheless, the highest contributions to phenotypic variance were RIL (14.2%) and RIL × test × environment (14.8%). However, H² of resistance to P. cubensis Clade 2 / mating type A1 was lower (0.32) than that for Clade 1 / mating type A2 (0.63), which predicts greater difficulty to introgress these new traits (Toporek et al. 2021).

qPcub-8.3 was responsible for 22% of phenotypic variation observed across all experiments, and qPcub-10.3-10.4 was responsible for 21.1% of the phenotypic variation, both with LOD scores above 9, making them the major QTL identified in this study. The major QTL previously identified in Toporek et al. (2021) for resistance to P. cubensis Clade 1 / mating type A2 were not identified in this study. Instead, the major QTL qPucb–8.3 (peak 81.7 cM; 1.5-Lod: 73.1–83.7 cM) was located upstream of qPcub-8.2 (peak: 91.6 cm; 1.5 LOD: 87.6-94.0 cM) and qPcub-10.3-10.4 (peak: 60.5 cM ; 1.5 LOD: 60.1-61.0 cM) was downstream of qPcub-10.1 (peak: 1.3 cM; 1.5 LOD: 0-5.0 cM). qPcub-10.3-10.4 does overlap with minor QTL qPcub-10.5, although their peaks are >7 cM away. qPcub-10.5 also appears to be the same minor QTL qPcub-10.2 observed previously (Toporek et al. 2021). The other minor QTL, qPcub-5.2 (peak: 81.4 cM; LOD 69.0-105.3 cM), found only across growth chamber studies, was also located far downstream of minor QTL qPcub-5.1 in Toporek et al. (2021) (peak: 0.5 cM; 1.5 LOD: 0-10.5 cM). Overall, chromosomes 8 and 10 appear to be important for CDM resistance in C. melo, with four major QTL located within them identified after screening with two genetically distinct P. cubensis isolates.

The major QTL qPcub-8.3 had a much larger 1.5 LOD interval and contained many more genes compared to qPcub-10.3-10.4. Within qPcub-8.3, MELO3C008820 and MELO3C008838, a putative disease resistance gene analog NBS-LRR and a receptor-like kinase, respectively, were identified. Within the much smaller range of qPcub10.3, a single receptor-like kinase was identified, MELO3C011814. The largest plant resistance (R) gene family encodes NBS-LRR type genes (McHale et al. 2006). Zhang et al. (2013) identified six NBS-LRR type genes within one major QTL for CDM resistance in C. sativus. In Arabidopsis, NBS-LRR genes RPP13 and RPP8 have been shown to confer downy mildew resistance (Bittner-Eddy et al. 2000; McDowell et al. 1998). RLKs are also a major class of plant R genes (Sekhwal et al. 2015). The RLK gene CsLRK10L2 was recently found in association with CDM resistance QTLs in C. sativus (Berg et al. 2020). No MLO resistance genes were identified in this study like those found within the most significant QTL in Toporek et al. (2021).

With the recent work by Purayannur et al. (2020) highlighting the differences in P. cubensis Clade 1 vs Clade 2 effectors and expression, it is not surprising that screening with two genetically distinct isolates, SS-01 and WAL-01, (Toporek et al. 2021) produced different reactions. In fact, isolate-dependent inheritance has been observed in C. sativus in a large study with 26 isolates of P. cubensis (Chen et al. 2020) and again with seven isolates (Katz et al. 2017), with the latter identifying multiple different QTLs when exposed to different isolates. Conversely, C. melo PI 124111F, derived from PI 124111 like MR-1, demonstrated broad-spectrum resistance to many P. cubensis isolates (Cohen and Eyal 1987). The analyses performed in this study and by Toporek et al. (2021) demonstrate that broad-spectrum CDM resistance may have different underlying mechanisms which are isolate-dependent.

This research identifies CDM resistance QTL in C. melo to an isolate genetically related to that of the 2004 epidemic population of P. cubensis in the United States. With the predominance of Clade 2, mating type A1 isolates in the eastern United States, these QTL will be the more important to incorporate. This study and the one by Toporek et al. (2021) represent the first analyses of the underlying resistance mechanisms of a single C. melo breeding line with two isolates of P. cubensis with highly different genetic backgrounds. A total of 31 KASP markers were developed across four major QTL. Future breeding lines incorporating resistance alleles from MR-1 can be rapidly screened for introgression of major QTL for CDM resistance using the KASP markers developed in this study.

Funding: This material is based upon work that is supported by the National Institute of Food and Agriculture, U. S. Dept. of Agriculture, under award numbers 2020-67034-31741 and project number SC-1700536. Technical Contribution No. 6XXX of the Clemson University Experiment Station.

Conflict of interest: The authors declare that they have no conflict of interest.

Ethics approval: Not applicable

Consent to participate: Not applicable

Consent for publication: Authors consent for publication

Availability of data and material: The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Code availability: Not applicable

Author contribution statement SMT, WPW, SEB, and APK designed and implemented the experiments. SMT phenotyped the population. SMT and SEB analyzed the data. SMT wrote the manuscript. All authors read and approved the final manuscript.

Acevedo-Garcia J, Kusch S, Panstruga R (2014) Magical mystery tour: MLO proteins in plant immunity and beyond. New Phytologist 204:273–281
Balass M, Cohen Y, Bar-Joseph M (1992) Identification of a constitutive 45 kDa soluble protein associated with resistance to downy mildew in muskmelon (Cucumis melo L.), line PI 124111 F. Physiol Mol Plant Pathol 41:387–396
Bates D, Mächler M, Bolker B, Walker S (2015) Fitting linear mixed-effects models using lme4. Journal of Statistical Software 67:1–48
Bittner-Eddy PD, Crute IR, Holub EB, Beynon JL (2000) RPP13 is a simple locus in Arabidopsis thaliana for alleles that specify downy mildew resistance to different avirulence determinants in Peronospora parasitica. Plant J 21: 177–188
Branham, SE, Levi A, Katawczik M, Fei Z, Wechter WP (2018) Construction of a genome-anchored, high-density genetic map for melon (Cucumis melo L.) and identification of Fusarium oxysporum f. sp. melonis race 1 resistance QTL. Theor. Appl. Genet. 131:829-837
Branham SE, Levi A, Katawczik M, Wecther WP (2019) QTL mapping of resistance to bacterial fruit blotch in Citrullus amarus. Theor Appl Genet. 132:1463–1471
Branham SE, Daley J, Levi A, Hassell R, Wechter WP (2020) QTL mapping and marker development for tolerance to sulfur phytotoxicity in melon (Cucumis melo). Front. Plant Sci. 11:1097
Broman KW, Wu H, Sen S, Churchill GA (2003) R/qtl: QTL mapping in experimental crosses. Bioinformatics 19:889–890
Broman KW, Sen S (2009) A guide to QTL mapping with R/qtl, vol 46. Springer, New York
Cao Y, Diao Q, Chen Y, Jin H, Zhang Y, and Zhang H (2021) Development of KASP markers and identification of a QTL underlying powdery mildew resistance in melon (Cucumis melo L.) by bulked segregant analysis and RNA-seq. Front Plant Sci. 11: 593207
Chen T, Katz D, Naim YB, Hammer R, Daniel BHB, Rubin AE, Cohen Y (2020) Isolate-dependent inheritance of resistance against Pseudoperonospora cubensis in cucumber. Agronomy 10:1086
Cohen Y (1981) Downy mildew of cucurbits. Pages 341-354. In: The downy mildews. D. M. Spencer, eds. Academic, London
Cohen Y, Eyal H (1987) Downy mildew-, powdery mildew- and fusarium wilt-resistant muskmelon breeding line PI-124111F. Phytoparasitica 15:187–195
Daley JD, Branham S, Levi A, Hassell R, Wechter P (2017) Mapping resistance to Alternaria cucumerina in Cucumis melo. Phytopathology 107:427–432
D’Arcangelo K, Rahman A, Miles TD, Quesada LM (2020) Utilizing a population genetics approach to provide crop-specific management strategies for cucurbit downy mildew. (Abstr.) Phytopathology 110:S1.7
Epinat C, Pitrat M (1989) Inheritance of resistance of three lines of muskmelon (Cucumis melo) to downy mildew (Pseudoperonospora cubensis). In: Thomas CE (ed) Evaluation and enhancement of cucurbit germplasm. Cucurbitaceae Proceedings, Charleston, SC, pp 133–135
Garcia-mas J, Benjak A, Sanseverino W, Bourgeois M, Mir G, González V M (2012) The genome of melon (Cucumis melo L.). Proc. Natl Acad. Sci. U.S.A.109:11872–11877
Goodwin, SB (1997) The population genetics of Phytophthora. Phytopathology 87:462-473
Haley CS, Knott SA (1992) A simple regression method for mapping quantitative trait loci in line crosses using flanking markers. Heredity 69:315–324
Holmes GJ, Ojiambo PS, Hausbeck MK, Quesda-Ocampo L, Keinath AP (2015) Resurgence of cucurbit downy mildew in the United States: a watershed event for research and extension. Plant Dis. 99:428–441
Katz D, Hammer R. Naim, YB, Cohen Y (2018) Resistance to downy mildew in cucumber breeding lines. Phytoparasitica 2018: 46, 320
Keinath, AP (2019) Integrated management of downy mildew on slicing cucumber with fungicides and host resistance but not trellising. Plant Dis. 103:2592-2598
Keinath, A. P., Miller, S. A., and Smart, C. D. 2019. Response of Pseudoperonospora cubensis to preventative fungicide applications varies by state and year. Plant Health Prog. 20:142-146
Kozik E, Klosinska U, Call A, Wehner T (2013) Heritability and genetic variance estimates for resistance to downy mildew in cucumber accession Ames 2354. Crop Sci 53:177–182
Li L, He H, Zou Z, Li Y (2018) QTL Analysis for downy mildew resistance in cucumber inbred line PI 197088. Plant Disease 102:1240–1245
Manichaikul A, Moon JY, Sen Ś et al (2009) A model selection approach for the identification of quantitative trait loci in experimental crosses, allowing epistasis. Genetics 181:1077–1086
McDowell JM, Dhandaydham M, Long TA, Aarts MGM, Goff S, Holub EB, Dangl JL (1998) Intragenic recombination and diversifying selection contribute to the evolution of downy mildew resistance at the RPP8 locus of Arabidopsis. Plant Cell 10:1861–1874
McHale L, Tan X, Koehl P, and Michelmore RW (2006) Plant NBS-LRR proteins: Adaptable guards. Genome Biol. 7:212
Nowicki N, Hadziabdic D, Trigiano RN, et al (2021) "Jumping Jack": Genomic microsatellites underscore the distinctiveness of closely related Pseudoperonospora cubensis and Pseudoperonospora humuli and provide new insights into their evolutionary past. Front Microbiol. 12:686759
Ojiambo PS, Gent DH, Quesada-Ocampo LM, Hausbeck MK, Holmes GJ (2015) Epidemiology and population biology of Pseudoperonospora cubensis: a model system for management of downy mildews. Annu. Rev. Phytopathol. 53:223–246
Olaya G, Kuhn P, Hert A, Holmes G, and Colucci S (2009) Fungicide resistance in cucurbit downy mildew. (Abstr.) Phytopathology 99:S171
Quesda-Ocampo LM, Granke LL, Olsen J, Gutting HC, Runge F, Thines M, Lebeda A, Hausbeck MK (2012) The genetic structure of Pseudoperonospora cubensis populations. Plant Dis. 96:1459–1470
Rahman A, Standish JR, D’Arcangelo KN, and Quesada-Ocampo LM (2021) Clade-specific biosurveillance of Pseudoperonospora cubensis using spore traps for precision disease management of cucurbit downy mildew. Phytopathology 111:312-320
R Core Team (2017) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria
Runge F, Choi Y, Thines M (2011) Phylogenetic investigations in the genus Pseudoperonospora reveal overlooked species and cyptic diversity in the P. cubensis species cluster. Eur J Plant Pathol. 129:125–146
Perchepied L, Bardin M, Dogimont C, Pitrat M (2005) Relationship between loci conferring downy mildew and powdery mildew resistance in melon assessed by quantitative trait loci mapping. Phytopathology 95:556–565
Purayannur S, Cano LM, Bowman M, Childs KL, Quesada-Ocampo LM (2020) Host-specific effectors of the cucurbit downy mildew pathogen Pseudoperonospora cubensis. (Abstr.) Phytopathology 110:S1.3
Sekhwal M, Li P, Lam I et al (2015) Disease resistance gene analogs (RGAs) in plants. Int J Mol Sci 16:19248–19290
Thomas CE (1986) Downy and powdery mildew resistant melon breeding line MR-1. HortScience 21:329
Thomas CE, Cohen Y, McCreight JD, Jourdain EL, Cohen S (1988) Inheritance of resistance to downy mildew in Cucumis melo. Plant Dis 72:33–35
Thomas A, Neufeld KN, Seebold KW, Braun CA, Schwarz MR, and Ojiambo PS (2018) Resistance to fluopicolide and propamocarb and baseline sensitivity to ethaboxam among isolates of Pseudoperonospora cubensis from the eastern United States. Plant Dis. 102:1619-1626
Toporek SM, Keinath AP (2021) Clade and mating type distribution and population structure of P. cubensis on Cucumis melo in the eastern United States. (Abstr.) Phytopathology 111:S2.35.
Toporek SM, Branham SE, Katawczik, ML, Keinath AP, Wechter WP (2021) QTL mapping of resistance to Pseudoperonospora cubensis Clade 1, Mating Type A2, in Cucumis melo. Theor. Appl. Genet. 134:2577–2586
Wallace EC, D’Arcangelo KN, Quesada-Ocampo, LM (2020) Population analyses reveal two host-adapted clades of Pseudoperonospora cubensis, the causal agent of cucurbit downy mildew, on commercial and wild cucurbits. Phytopathology 110:1578-1587
Wang K, Li M, Hakonarson H (2010) ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res. 38:e164e164
Zhang SP, Liu MM, Miao H, Zhang SQ, Yang YH, Xie BY, Wehner TC, Gu XF (2013) Chromsomal mapping and QTL analysis of resistance to downy mildew in Cucumis sativus. Plant Dis. 97:245-251

Table 1 Variance components of percent diseased leaf area of a Cucumis melo population in response to inoculation with Pseudoperonospora cubensis Clade 2, mating type A1

Variance Component	Variance	%V_P^a	Standard Dev
RIL (R)	25.1	14.21	5.0
Test (T)	7.5	4.22	2.7
Environment (E)	5.8	3.31	2.4
T x E	0.02	0.01	0.04
R x E	8.2	4.61	2.9
R x T	1.7	0.98	1.3
R x T x E	26.2	14.81	5.1
Block within T	0.3	0.17	0.6
Residual	101.9	57.67	10.1

^a Percentage of the total phenotypic variance

Table 2 Quantitative trait loci (QTL) associated with percentage of diseased leaf area averaged across and within tests after artificial inoculation with Pseudoperonospora cubensis Clade 2, mating type A1

QTL^a	Dataset mean	Peak (cM)	1.5-LOD interval (cM)	Range (Mb)^b	LOD	Vp(%)^c	P value	Additive	Genes
qPcub-5.2	Across Growth Chamber	81.4	69.0–105.3	16.3–24.7	3.3	15.71	1.1 × 10^-4	2.4	285
*qPcub-8.3*	Greenhouse 1	82.8	74.5–84.5	9.2–25.4	8.8	17.9	3.4 × 10^-10	4.3	616
	Growth Chamber 2	82.8	69.2–94.0	8.2–29.4	4.0	11.3	2.2 × 10^-5	3.3	919
	Across Greenhouse	82.5	74.3–84.6	8.9–25.4	9.5	18.5	7.4 × 10^-11	3.1	646
	Across Growth Chamber	77.3	72.1–83.4	8.5–24.9	6.1	15.6	1.8 × 10^-7	3.3	651
	Across All	81.7	73.1–83.7	8.6–25.1	9.1	22.0	1.6 × 10^-10	3.0	655
*qPcub-10.3*	Greenhouse 1	60.1	59.8–60.4	4.6–4.7	9.0	18.2	2.5 × 10^-10	-26.6	7
	Across Greenhouse	60.1	59.8–60.4	4.6–4.7	9.4	18.4	8.3 × 10^-11	-18.9	7
*qPcub-10.4*	Greenhouse 1	60.5	60.1–61.0	4.6–4.7	10.0	20.6	2.4 × 10^-11	28.4	16
	Across Greenhouse	60.5	60.1–61.0	4.6–4.7	10.6	21.1	5.3 × 10^-12	20.3	16
qPcub-10.5	Across All	67.0	60.1–81.4	4.5–6.3	3.7	8.3	4.4 × 10^-5	1.9	147

^a Bolded QTL were validated with non-parametric testing

^b Physical region of the genome corresponding to the 1.5-LOD interval of the QTL

^c Percent of the phenotypic variation explained by the QTL

^d Additive effect of the QT

Download PDF

Journal Publication

published 03 Apr, 2023

Read the published version in Theoretical and Applied Genetics →

Reviewers agreed at journal
12 Sep, 2022
Reviewers invited by journal
21 Jul, 2022
Editor assigned by journal
09 Jun, 2022
First submitted to journal
08 Jun, 2022

You are reading this latest preprint version

QTL Mapping of Resistance to Pseudoperonospora Cubensis Clade 2, Mating Type A1, in Cucumis Melo and Dual-clade Marker Development

Status:

Journal Publication

Version 1

Abstract

Figures

Key Message

Introduction

Materials And Methods

Results

Discussion

Declarations

References

Tables

Status:

Journal Publication

Version 1