Figures
Abstract
Powdery mildew is one of the most common fungal diseases in the world. This disease frequently affects melon (Cucumis melo L.) and other Cucurbitaceous family crops in both open field and greenhouse cultivation. One of the goals of genomics is to identify the polymorphic loci responsible for variation in phenotypic traits. In this study, powdery mildew disease assessment scores were calculated for four melon accessions, ‘SCNU1154’, ‘Edisto47’, ‘MR-1’, and ‘PMR5’. To investigate the genetic variation of these accessions, whole genome re-sequencing using the Illumina HiSeq 2000 platform was performed. A total of 754,759,704 quality-filtered reads were generated, with an average of 82.64% coverage relative to the reference genome. Comparisons of the sequences for the melon accessions revealed around 7.4 million single nucleotide polymorphisms (SNPs), 1.9 million InDels, and 182,398 putative structural variations (SVs). Functional enrichment analysis of detected variations classified them into biological process, cellular component and molecular function categories. Further, a disease-associated QTL map was constructed for 390 SNPs and 45 InDels identified as related to defense-response genes. Among them 112 SNPs and 12 InDels were observed in powdery mildew responsive chromosomes. Accordingly, this whole genome re-sequencing study identified SNPs and InDels associated with defense genes that will serve as candidate polymorphisms in the search for sources of resistance against powdery mildew disease and could accelerate marker-assisted breeding in melon.
Citation: Natarajan S, Kim H-T, Thamilarasan SK, Veerappan K, Park J-I, Nou I-S (2016) Whole Genome Re-Sequencing and Characterization of Powdery Mildew Disease-Associated Allelic Variation in Melon. PLoS ONE 11(6): e0157524. https://doi.org/10.1371/journal.pone.0157524
Editor: David Gent, US Department of Agriculture, UNITED STATES
Received: February 13, 2016; Accepted: June 1, 2016; Published: June 16, 2016
Copyright: © 2016 Natarajan et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files. Also all raw files are available from the NCBI database (accession number PRJNA300827). These information's were mentioned in "Materials and Methods" section and contact to corresponding author if further details required.
Funding: This research was supported by Golden Seed Project (Center for Horticultural Seed Development, Ministry of Agriculture, Food and Rural Affairs (MAFRA), Ministry of Oceans and Fisheries (MOF), Rural Development Administration (RDA) and Korea Forest Service (KFS): 213003-04-4-SB110. Export Promotion Technology Development Program (Grant No. 312065-05-4-HD030), Ministry of Agriculture, Food and Rural Affairs, Republic of Korea.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Melon (Cucumis melo L.) is a highly diversified eudicot diploid species (2n = 2x = 24) from the Cucurbitaceae family, which includes cucumber, watermelon, and squash. Melon was previously classified into two subspecies, C. melo ssp. agrestis and C. melo ssp. melo, which were differentiated based on pubescence on the hypanthium [1] and later divided into several edible and wild varieties [2]. Melon is an important vegetable/fruit crop cultivated worldwide and highly valued for its fruit quality. More than 31 million tons of melons were produced worldwide in 2012 (Food and Agriculture Organization, http://faostat.fao.org). Formerly, Africa was designated as the primary site of origin for melons; however, current data indicate that cucumber and melon may have originated in Asian regions [3]. In recent decades, melons have emerged as an attractive model system for observable biological traits, such as fruit ripening [4], plant sex determination [5], and phloem transport [6]. A rich diversity of genomic tools have been developed for melon including a physical map [7], genetic maps [8], an EST database (http://www.icugi.org) [9], reverse genetic tools [10], BAC sequences [11], and a microarray [12].
Melon (Cucumis melo L.) is susceptible to various biotic and abiotic stresses [12, 13], especially powdery mildew [14, 15], which is poses a problem for Cucurbitaceous crops in the open field as well as in greenhouses. The fungi Podosphaera xanthii (syn. Sphaerotheca fuliginea) and Golovinomyces cichoracearum (syn. Erysiphe cichoracearum) are the main causal agents of powdery mildew (PM), which impedes plant growth and limits fruit quality production worldwide [15]. Powdery mildew is caused by a range of fungal races, and despite its severity on cucurbits, various accessions or cultivars show resistance to specific races [16]. However, commercial development of resistance genes against PM pathogens is hampered due to limited knowledge of melon genetic variation [17]. Analysis of a large number of races identified Go and Px resistance sources [18]; however, more information is required about the underlying genetic basis of the resistance, including the molecular underpinnings. Important resistance genes including Go and Px are listed in the Cucurbit Genetics Cooperative (cgc.ncsu.edu) for Cucurbitaceae species [19]. In addition, molecular markers based on restriction fragment length polymorphism (RFLP), random amplified polymorphic DNA (RAPD) [20], amplified fragment length polymorphism (AFLP) [21], single nucleotide polymorphism (SNP) [22] and simple sequence repeat (SSR) [23] have been implemented to study the genetic diversity of melon.
Next generation sequencing (NGS) technology has been used to extend our insights in genomic research for developing molecular markers, identification of genetic variation and gene discovery using sequencing approaches [24]. Among these technologies, whole-genome re-sequencing (WGR) has been successfully utilized to study genome diversity and discover sequence variations in crops including rice, maize, pigeon pea, chickpea, soybean, and common bean [24]. In addition to identification of genetic polymorphisms such as single nucleotide polymorphism (SNP) and insertion/deletion polymorphism (InDel), WGR permits the detection of copy number of variation (CNV) and presence/absence variation (PAV) [25]. Based on Illumina technologies, genome sequences of melon [26], watermelon [27], and cucumber [28] have been published. The melon genome was recently sequenced using the double-haploid line DHL92, derived from a cross between the melon cultivars of PI 161375 (Songwhan Charmi) and T111 (Piel de Sapo). The resulting assembled genome comprised 27,427 annotated genes with 17% of the genome being transposable elements (TEs) [26]. Considering the advantages of NGS, we have chosen four melon accessions for re-sequencing to characterize their genotypic variation in terms of SNPs, InDels, and structure variations (SVs). In addition, QTLs associated with disease resistance genes were detected for the re-sequenced variants based on available R genes from the reference genome.
Materials and Methods
Plant materials and disease assessment
A total of four melon (Cucumis melo L.) accessions, ‘SCNU1154’, ‘Edisto47’, ‘MR-1’, and ‘PMR5’, were included in this study based on PM races. Collection of PM strains, characterization and inoculation methods were followed from our published paper [29]. The foliage leaves were harvested from glasshouse-grown plants, cut into pieces and placed on MS medium in Petri dishes. For inoculation, plants were exposed to powdery mildew by conidial suspension/spraying [30] or by placement into a sedimentation tower [31]. Inoculated samples were incubated at 26°C on a 16-h light and 8-h dark cycle for two weeks. Powdery mildew disease severity was accessed using the Wilcoxon t-test with Bonferroni correction, where lesion area infection covered up to 10% (-), 10% to 30% (+), or more than 31% (++) on each sample. Additionally each melon accession was classified as resistant or susceptible using a disease index (DI) scale (0 to 5) based on the previously described calculation [32].
DNA isolation and re-sequencing
Young leaves from each melon line were harvested from two-week-old seedlings and then subjected to genomic DNA extraction using the Qiagen DNeasy Plant Mini Kit. DNA was randomly fragmented, followed by adapter ligation and then band size-fractionated by gel electrophoresis. A sequence library of DNA of the desired length was constructed for each melon line according to the manufacturer’s instructions (Illumina, Inc., San Diego, CA). Whole-genome re-sequencing was performed using the Illumina Hiseq 2000 platform with the constructed paired-end (PE) sequence libraries at the 380Beijing Genomics Institutes (BGI)-Shenzhen (Shenzhen, China).
Detection of SNPs and InDels
The recently released genome of the melon (Cucumis melo L.) accession DHL92 was downloaded from Melonomics (http://melonomics.net, V3.5) [26] and used as a reference genome. The sequencing data generated in this work for each melon line were examined for quality assessment using FASTX-toolkit (http://hannonlab.cshl.edu/fastx_toolkit/). The resulting clean reads from the PE sequencing for each melon accession were mapped to the melon reference genome using the Burrows-Wheeler Aligner (BWA) algorithm (v0.7.7) with default parameters [33]. Further, using SAMtools (V0.1.19) the resulting Binary Alignment/Map (BAM) and Sequence Alignment/Map (SAM) files were converted for sorting and indexing alignments [33]. Mark duplicates in Picard tools (V1.106) was used to discard duplicates, and the final sorted bam results were used for downstream analysis. In addition, detection of SNPs was carried out employing the multi-sample SNP calling approach using the Genome Analysis Toolkit (GATK, V3.1) [34]. This program performs local re-alignment and re-calibration to generate putative SNPs and InDels for each sample with the output format of VarScan2 (VCF). The output file consists of information related to variants and their positions along with SNP quality of coverage, mapping, genotype, and read-depth statistics [35].
SNP and InDel annotation
From the VCF files obtained through filtering, SNPs and InDels were annotated based on genomic location and classified by the likely effects of the variations including functional categories using SnpEff (V4.11) [36]. The SnpEff binary database file (.bin) was generated using the melon genome annotation file (gff3) and the genome sequence. This generated database was used to annotate the effects of SNP by region, effect (high, moderate, low and modifier), and functional class (missense, nonsense, and silent) for all individuals. The generated output files, both HTML and text files, were subjected to further analysis including for localization of coding, non-coding, intronic, exonic, start codon, stop codon, upstream/downstream, splice sites, 5’ UTR and 3’ UTR regions. Then, functional enrichment analysis based on gene ontology (GO) was carried out for genes with detected SNP/InDel allelic variations. Blast2Go (https://www.blast2go.com) was also used for functional analysis of each variant; GO terms were determined with Fisher’s Exact Test and divided into three categories: biological process (BP), cellular component (CC), and molecular function (MF) [37]. Finally, detected genes were mapped to the KEGG (Kyoto Encyclopedia of Genes and Genomes) [38] database for analyzing biological pathways, and protein family annotations were assigned according to Pfam (http://pfam.xfam.org/) [39].
Variations detected in disease resistance genes
In this study, different classes of resistance (R) genes were collected from the PRGdb database (Plant Resistance gene Database, http://prgdb.crg.eu/) [40] and taken from reports of resistance genes involved in melon disease [19]. The putative functions of genes with detected variations were determined through BLAST search using Blast2GO. This program allows the examination of putative functions of genes and characterization based on gene ontology. Further, domain analysis was carried out using the Pfam database and variant effects were correlated according to detected reads. The physical map of detected variants related to disease responses was constructed using MapChart (v2.2) [41].
Detection of structure variations (SVs)
Abnormal orientation of paired-end reads aligned using the reference genome was analyzed using BreakDancer (v1.1.2). This program detects structural variations by clustering and comparing with previously defined SV types [42]. The structural variations of deletions (DEL), insertions (INS), inversions (INV), inter-chromosomal translocations (CTX), and intra-chromosomal translocations (ITX) were identified using the default parameters.
Validation of SNPs in melon accessions
A total of 12 SNPs (~ 4 from each PM responsive chromosomes of 2, 5, and 12) with their corresponding 500 base pair flanking sequences were selected, and primer pairs were designed for experimental verification. The presence of each SNP was confirmed in melon accessions (SCNU1154: susceptible; Edisto, MR-1, and PMR5: resistant). The designed forward and reverse primer pairs were used for PCR amplification to verify variants. PCR amplifications were carried out using a Takara PCR Thermal Cycler with reaction mixtures (total volume of 20 μL) containing 1 × buffer, 2.0 mM MgCl2, 0.2 mM dNTPs, 0.2μM primers, 5 ng template DNA and 0.5 units Taq polymerase (PROMEGA, Madison, USA). The PCR cycle conditions included initial denaturation at 94°C for 5 min, then 30 cycles of denaturation at 94°C for 30 s, annealing at 59°C for 30s, extension at 72°C for 30s, and a final elongation at 72°C for 5 min. Further, PCR products were visualized using agarose gel electrophoresis and purified PCR fragments were used for sequencing (Macrogen, Korea).
Data availability
The identified SNPs and InDels for each individual line were compared between samples for identification of unique and shared variations. The identified variations were used to plot a genome-wide chromosomal representation of structural variations using the Circos program [43]. The generated paired-end sequences were deposited into NCBI BioProject database (accession number—PRJNA300827, http://www.ncbi.nlm.nih.gov/bioproject/).
Results and Discussion
Disease assessment
In recent decades, a number of races have been analyzed to identify the PM inheritance for melon accessions/cultivars. Environmental factors such as temperature between 25–30°C and 98% relative humidity are requirements for sporulation of PM pathogens and severely affect the chlorotic leaves of diseased plants. The present study was carried out to confirm the PM response in four different melon accessions of SCNU1154, Edisto47, MR-1, and PMR5 by disease assessment. Young leaves were inoculated with Podosphaera xanthii (PM pathogen) and disease severity was measured for each line using disease index scores (0 to 5). Further confirmation of disease severity was accessed using Wilcoxon t-test with Bonferroni correction, where a visual cutoff range was fixed based on lesion area infection, 10% (-), 10% to 30% (+), and more than 31% (++) on each sample (S1 Fig). According to bioassay results on melon leaves infected with PM, ‘SCNU1154’ was highly susceptible (S, DI = 5), and Edisto47, MR-1, PMR5 lines were highly resistant (R, DI = 0). In addition, earlier work with other races also confirms that Edisto47, MR-1, PMR5 lines are highly resistant to Podosphaera xanthii in melon [44].
Whole-genome sequencing
The four melon accessions ‘SCNU1154’, ‘Edisto47’, MR-1, and ‘PMR5’ were successfully sequenced using Illumina Hiseq2000 sequencing technology. Four individual paired-end DNA libraries were constructed and yielded 754,759,704 reads with an average of 15.62 million reads per accession. In total, 77.54% to 86.56% of reads were aligned on the reference melon genome. On average, each melon accession covered 82.64% of the reference genome, suggesting that our re-sequenced genomes are closely related to the reference genome. The statistics of generated raw reads, mapped alignments, and coverage of each line are summarized in Table 1. An average of 37.15% GC content was estimated for the re-sequenced samples. Mapped sequences were used for subsequent analyses.
Detection and characterization of SNPs and InDels
Comparison to the reference genome to detect SNP variations were carried out separately for each accession using the SAMtools software. An average of 1,851,346 SNPs were detected in the accessions and the SNPs detected in ‘SCNU1154’, ‘Edisto47’, ‘MR-1’, and ‘PMR5’, respectively were shown in Table 2. These rates of SNP density in the genome correspond to approximately one change in every 217 bases (SCNU1154), 264 bases (Edisto47), 199 bases (MR-1), and 207 bases (PMR5). The identified SNPs in ‘SCNU1154’ comprised 69.02% transitions and 30.97% transversions. Among these, 6.96% were heterozygous and 93.03% were homozygous. The obtained SNPs from ‘Edisto47’ were 66.84% transitions and 33.13% transversions. Approximately 5.2% were heterozygous and 94.74% were homozygous. The SNPs of ‘MR-1’ comprised 66.85% transitions and 33.14% transversions, with approximately 5.43% heterozygous and 94.56% homozygous. The detected SNPs of ‘PMR5’ were 66.46% transitions and 33.53% transversions, with 6.62% heterozygous and 93.37% homozygous (S1 Table). In addition, transition/transversion ratio variations were observed, with average of 2.06 for all accessions (Fig 1). Furthermore, a total of 532,910, 420,691, 449,256, and 574,839 InDels were identified in ‘SCNU1154’, ‘Edisto47’, ‘MR-1’, and ‘PMR5’, respectively, by comparing their individual genome sequences to the reference genome (Table 2). Among these detected InDels, ‘SCNU1154’ had 178,041 insertions and 354,862 deletions; ‘Edisto47’ had 131,313 insertions and 289,374 deletions; ‘MR-1’ had 175,534 insertions and 273,722 deletions; and ‘PMR5’ had 175,891 insertions and 398,944 deletions (S2 Fig). The approximate InDel density variant rate observed for each accession was one change in every 763 bases (SCNU1154), 967 bases (Edisto47), 905 bases (MR-1), and 707 bases (PMR5).
Chromosomal distribution of SNPs and InDels
The reference genome of melon, including Pseudo chromosomes 1 to 12 (Chr.1 to 12) as well as non-anchored scaffolds and contigs (Chr.0) (hereafter pseudo chromosomes are referred to as chromosomes in this manuscript) were used for analyzing chromosomal distributions of the identified SNPs. The SNPs in all accessions were widely distributed to each chromosome including non-anchored scaffolds and contigs (S3A Fig). Chr.0 contained the fewest SNPs, with an average of 58,117 among all accessions. SNP coverage in ‘SCNU1154’ was lowest in Chr.2 and Chr.9 compared to other chromosomes. Similarly, Chr. 1, 2, and 9 from ‘Edisto47’ and Chr. 9 from ‘PMR5’ had fewer SNPs compared to other chromosomes within the respective accessions. The ‘MR-1’ SNPs were distributed among all chromosomes. The length of InDels (insertions and deletions) ranged between 1 to 22 bp according to accession (Fig 2). In this study, the largest InDel size (22 bp, 3 occurrences) was detected in ‘PMR5’ and most of the InDels were in less than 10 bp. In addition, InDel lengths varied between the accessions (S2 Table). Further, a high rate of single bp InDels existed in all accessions among the detected InDels. The distributions of identified InDels in the melon chromosomes is shown in S3B Fig. Comparing all accessions, the highest abundance of InDels was detected in Chr.1 and a small proportion of InDels was detected in Chr.0 (11,297 avg.). The remaining chromosomes showed similar patterns of InDel distribution according to total detected InDels per accession. The short InDels (1–5 base pairs) may have deleterious effects on gene functionality during expression or transcriptional processes [25]. Overall, the maximum number of detected SNPs and InDels was in ‘PMR5’, with relatively low variation observed in ‘Edisto47’. The frequency of SNPs was highest in ‘MR-1’ in comparison to the other accessions, while the most InDels were detected in ‘PMR5’. The detected SNPs and InDels from each accession were visualized in the melon chromosomes using Circos (Fig 3).
Distribution of (A) SNPs and (B) InDels are represented in each melon pseudo-chromosome.
Annotation of SNPs and InDels
To annotate the detected SNPs and InDels from SAMtools, sequence information and annotation files of the melon genome were used to evaluate the possible effects of the variants in genomic, exonic, and functional categories. Approximately 46.5%, 42.7%, 43.7%, and 44.8% SNPs were in intergenic regions, 5.7%, 6.7%, 6.2%, and 6% were intronic, and 1.7%, 1.7%, 1.6%, and 1.6% exonic in SCNU1154’, ‘Edisto47’, ‘MR-1’ and ‘PMR5 accessions, respectively. The effects of SNP variations were classified into four classes, “high effect”, “moderate effect”, “low effect”, and “modifier”, according to SnpEff [36]. In all, 98% (Avg.) of SNPs were classified as sequence modifiers (introns or affecting non-coding genes) and moderate effect accounted for 0.81% of the SNPs on average. The low effect variants ranged from 0.95 to 1.08% according to accession. The remaining high effect variants ranged from 950 to 1,192 SNPs (0.02% avg.) within the tested accessions. S3 Table describes the proportion of variants in each effect category.
Functional classes of missense, nonsense, and silent variants were evaluated for the detected SNPs. A total of 1,192 SNPs (0.034% of total SNPs) were considered as high effect variants in ‘SCNU1154’ and included 49.05% missense SNPs and 49.87% silent SNPs, with a 0.98 missense/silent ratio. In ‘Edisto47’, 950 SNPs were assigned as high effect (0.031% of total SNPs) and comprised 47.8% missense SNPs and51.26% silent SNPs, with a missense/silent ratio of 0.93. In the case of ‘MR-1’, a total of 1,176 SNPs (0.029% of total SNPs) were measured as high effect variants, with 48.049% missense SNPs and 50.99% silent SNPs and a 0.94 missense/silent ratio. The ‘PMR5’ accession exhibited 1,186 SNPs (0.031% of total SNPs) in the high effect variation category; 49.30% were missense and 49.63% were silent. In addition, small proportions of nonsense SNPs were detected in SCNU1154’,’Edisto47’,’MR-1’and ‘PMR5 accessions (1.07%, 0.93%, 0.96%, and 1.06%, respectively; Fig 4).
Genetic variation of SNPs detected in (A) SCNU1154, (B) Edisto47, (C) MR-1, and (D) PMR5.
Similar methodology was used to annotate the detected InDels based on the melon reference genome using SnpEff. Among total InDels, 35.1%, 32.9%, 39.2%, and 28.3% were in intergenic regions, 5.6%, 5.8%, 7.8%, and 9% were intronic, and 0.2%, 0.2%, 0.3%, 0.2% were exonic in ‘SCNU1154’, ‘Edisto47’, ‘MR-1’, and ‘PMR5’, respectively. An average of 99.5% InDels (among all InDels from all accessions) was classified as sequence modifiers and 0.12% as moderate effect. The low effect variants were observed in the range from 0.09 to 1.15% according to accession. The remaining high effect variants accounted for 0.19% in total (S3 Table). The percentages of high-effect InDels (SCNU1154: 0.18%, Edisto47: 0.18%, MR-1: 0.23% and PMR5: 0.15%) were fairly small. In addition, possible functional roles such as frameshift, in-frame, and splice-site variants were assessed for all detected InDels (Fig 5). The observed InDels with high effect variants were higher than for SNPs, likely because InDels may cause frameshift effects in the sequence [45].
Genetic variation of InDels detected in (A) SCNU1154, (B) Edisto47, (C) MR-1, and (D) PMR5.
Comparative analysis and annotation of variant genes
Detected SNPs and InDels were examined to identify non-redundant genes in each accession. This analysis revealed that 24,197 genes with detected SNPs and 26,453 genes with detected InDels were in ‘SCNU1154’, 26,063 genes with detected SNPs and 26,166 genes with detected InDels were in ‘Edisto47’, 26,693 genes with detected SNPs and 26,348 genes with detected InDels were in ‘MR-1’, and 26,532 genes with detected SNPs and 26,453 genes with detected InDels were in ‘PMR5’. These genes were compared to each other to identify the unique and common genes containing SNP and InDel variations between all accessions (Fig 6). This comparative result revealed the presence of 71 unique SNP-containing genes in ‘SCNU1154’, possessing 1,371 SNPs. Likewise, 23, 136, and 19 unique genes with 823, 19,931, and 1,298 SNPs were found in ‘Edisto47’, ‘MR-1’, and ‘PMR5’, respectively. A total of 74, 22, 68, and 50 unique genes were identified with 613, 156, 1,438, and 1,298 InDel variations in ‘SCNU1154, ‘Edisto47’, ‘MR-1’, and ‘PMR-5’ accessions, respectively. Notably, 23,331 unique SNP genes and 25,896 unique InDel genes were identified as common to all four accessions.
Comparison of (A) SNP and (B) InDel variations detected in each melon accession.
We used the Blast2GO program for functional enrichment analysis of detected high and moderate effect variants (SNPs and InDels) in all accessions. The closely related species, Cucumis sativus was used to obtain the putative functions using BLAST searches. Based on functional GO annotations of SNP-containing genes, the assigned GO terms belonging to the biological process class ranked the highest (avg. of 6,646,37% and 11, 3295,40%), followed by the molecular function class (avg. 5,845,32% and 85,473,30%), and the cellular component class (avg. 5,636,31% and 83,929,30%) for high- and moderate-effect variants, respectively. The GO terms obtained for InDel genes revealed that an average of 19,310 and 11,654 (40% and 39%) were in biological process, followed by cellular component with 14,423 and 11,654 (40% and 31%), and molecular function with 14,423 and 9,177 (30%, and 30%), respectively. Of note, ‘catalytic activity’ (GO: 0003824) and ‘binding’ (GO: 0005488) were highly represented GO terms in the molecular function category. Under the classification of biological process ‘cellular process’ (GO: 0009987) and ‘metabolic process’ (GO: 0008152) were significantly represented among GO terms. ‘Cell part’ (GO: 0044464) and ‘cell’ (GO: 0005623) are most abundant GO terms in cellular component (S4 and S5 Figs).
Structural Variants (SVs)
Variations such as deletions (DEL), insertions (INS), inversions (INV), inter-chromosomal translocations (CTX), and intra-chromosomal translocations (ITX) are defined as genomic structural variations if they consist of greater than 1 kb sequence rearrangement [46]. In the present study, SVs were detected for each melon accession by making use of the BreakDancer program with default parameters. We found a total of 47,743 (SCNU1154), 44,510 (Edisto47), 40,908 (MR-1), and 49,237 (PMR5) SVs with an average of 45.5% variation when compared to the reference genome (Table 3). For these SVs, InDels of deletion events (average = 45%) outnumbered insertions (average = 11%). The number of SVs found in this study was consistent with those found in other plant species [46]. The other variations including inversions (INV), inter-chromosomal translocations (CTX), and intra-chromosomal translocations (ITX) accounted for an average of 11%, 27.25% and 6.5%, respectively. In brief, SVs of individual melon accessions ‘SCNU1154’, ‘ Edisto47’, ‘MR-1’, and ‘PMR5’ were 53%, 58%, 56%, and 53% InDels, 12%, 10%, 11%, and 11% INV, 7%, 6%, 7%, and 6% ITX, and 28%, 26%, 26%, and 29% CTX, respectively. In addition, a total of 26,450 InDel, 5,339 INV, 14,234 CTX, and 3,214 ITX SVs detected in Chr.0, and CTX events were not detected in Chr.12 for any of the accessions (S4 Table).
Variants in resistance genes
The defense responses genes related to disease resistance were identified through functional annotations of detected SNPs and InDels compared to the reference genome. Previously reported disease resistant genes/QTLs from PRGdb in Cucumis melo were taken into consideration. A total of 411 putative disease resistance genes have been reported in melon, and this number is low compared to the numbers of resistance genes known in Arabidopsis (526), grape (690), and rice (1206) [26]. In melon, resistance genes have been identified for diseases including zucchini yellow mosaic virus (ZYMV) (Zym; [47]), watermelon mosaic virus-Morocco (Potyvirus) (Nm; [48]), aphis gossypii (Fn; [49]), papaya ring spot virus- watermelon type (PRSV) (Prv; [47, 50]), cucurbit aphid borne yellows virus (CABYV) (cab-1 and cab-2; [51]), beet pseudo yellows virus (BPYV; [52]), cucurbit yellow stunting disorder virus (CYSDV; [53]), lettuce infectious yellows virus (LIYV; [54]), cucurbit leaf crumple virus (CuLCrV; [55]), melon necrotic spot virus (MNSV; [56]), cucumber green mottle mosaic virus (CGMMV) (cgmmv-1 and cgmmv-2; [57]), and fusarium wilt (Fom-1; [58], Fom-2; [59, 60]). In addition, several powdery mildew resistance (Pm) genes have been mapped or cloned in the melon genome (Pm-1; [61], Pm-2; [62], Pm-3; [63], Pm-4 and Pm-5; [64], Pm-6; [65], Pm-8; [66], and Pm-F & Pm-G; [67]). In addition, several studies were reported that Mildew Locus O (MLO) genes act as susceptibility factors in PM disease and their inactivation of specific MLO genes (knock-out or knock-down) leads to mediating form of mlo resistance [68].
In this study, comprehensive domain analyses of encoded defense genes were performed. A total of 11,867 SNPs were identified in 585 defense-related genes in all four accessions and found to be distributed across all chromosomes. Among them, different classes of disease resistance-related protein domains were found, including CNL [CC-NB-LRR, proteins containing coiled-coil (CC), nucleotide-binding site (NBS), and leucine-rich repeat (LRR) domains], TNL [TIR-NB-LRR, Toll-interleukin receptor-like, nucleotide binding site (NB) and LRR domains], RLP (receptor serine/threonine kinase-like domain), and an extracellular LRR (ser/thr-LRR); in addition, other kinase domains with an extracellular LRR (RLK) were detected in all four accessions. The details of defense-related genes with detected SNPs are tabulated in S5 Table. Among 11,867 SNPs in defense-related genes, 4 SNPs predicted to be high effect variants (affecting splice site regions, start and stop codons) and 386 predicted as moderate effect variants (non-synonymous coding, codon insertions and deletions) were found within 139 defense-related genes. In addition, functional enrichment analysis showed that GO terms of catalytic activity (GO: 0003824), and binding (GO: 0005488) were highly represented in the molecular function category for these identified disease-related genes. Of note, these domains are commonly involved in plant defense mechanisms and mediate disease resistance in plants [40]. Similar to SNPs, 3,429 InDels were identified with 31 InDels (30 genes) and 14 InDels predicted to have high and moderate effects, respectively (S6 Table). Using the advances of NGS technology in melon re-sequencing, the presence and absence of defense-related genes has been demonstrated [69]. Leida et al [70] re-sequenced 175 melon accessions to identify the variability of SNPs and characterized sugar accumulation and fruit ripening-related genes. However, there has been no attempt reported for the development of QTLs associated with disease resistance-related genes in melon, especially for powdery mildew disease assessment using re-sequencing. Accordingly, we constructed a disease-associated QTL map for 390 SNPs and 45 InDels identified from this study as related to defense-response genes (Fig 7).
A total of 390 SNPs and 45 InDels are represented with physical map position. The different colors represent moderate-effect SNPs (black), high-effect SNPs (black with underline), moderate-effect InDels (red), and high-effect InDels (red with underline).
Recent evidences reveled that MLO gene specific knock-out or knock-down leads to PM resistant in arabidopsis, tomato, pea, pepper, grapevine and wheat [68]. For this study we have collected 14 MLO genes information from our published paper [71]. Further, the genes were characterized using basic bioinformatics analysis for identification of gene position, locus, and lengths. Then these informations were correlated with the present study and investigated for their SNPs and InDels variation. The comparative analysis of the MLO genes revealed 1,123, 1,123, 1,615, and 1,663 SNPs and 486, 386, 411, 464 InDels in ‘SCNU1154, ‘Edisto47’, ‘MR-1’, and ‘PMR-5’ accessions, respectively. Notably, 438 SNPs and 125 InDels were identified as common to all four accessions (S6 Fig and S8 Table). We hope that these MLO genes associated SNPs and InDels would help to accelerate the development of disease resistant marker against PM.
SNP validation in powdery mildew responsive chromosomes
In melon, several PM resistant genes and resistant trait QTL have been reported in chromosome 2, 5, and 12 [72]. We found 23, 24, 65 SNPs and 4, 5, 3 InDels from PM responsive chromosomes of 2, 5, and 12 respectively. Further, we conducted experimental validation for several detected SNPs related to possible disease resistance genes using 12 primer pairs (S7 Table) targeting of 12 SNPs from PM responsive chromosomes. During the process of retrieving the sequence for SNP selection we noticed poly N repeats those SNPs were omitted from further analysis. This inconsistency is due to the incompleteness of draft reference genome [73]. The selected SNPs were confirmed by their amplification by using four melon accessions of SCNU1154 (susceptible), Edisto (resistant), MR-1(resistant), and PMR5 (resistant). The SNPs were amplified by flanking primers in all four re-sequenced melon lines. In addition, MELO3C015353 and MELO3C015354 gene from chr.2 has been reported for PM candidate resistant gene in Edisto47 [72]. In this study, we observed a total of 4 SNPs in MELO3C015353 and MELO3C015354 genes from re-sequenced melon lines (S5 Table). MELO3C015354 is a CNL class of gene (CC-NB-LRR domains) and MELO3C015353 is a NL class gene lacking CC domain. The SNP (nucleotide transition of A to C—Chr.2: 994117) observed in MELO3C015354 gene caused a stop lost/splice region variant. Three missense non-synonymous variant was observed in MELO3C015353 gene (Chr.2: 985472, 986247, and 986300). Further, melon accession/allele-specific marker development are in progress form detected SNPs with associated genes. The identified SNPs associated with PM disease resistant QTL from re-sequenced melon lines may provide novel insights to improve the disease resistance in melon accessions by using advanced molecular technologies.
Conclusions
The importance of next generation technology is rapidly increasing in plant research by offering a variety of applications related to understanding genetic variation and facilitating marker identification/characterization in different plant varieties. In this present study, we have conducted PM disease assessment for ‘SCNU1154’, ‘Edisto47’, ‘MR-1’, and ‘PMR5’ accessions of melon and conducted whole-genome re-sequencing for each accession. Our large-scale cataloging of variations found among all accessions will be useful for marker development including SNP array design and fine mapping. In addition, the reference genome used in this study has gaps and uncertainty in the few regions of sequences. Particularly González VM et.al reported a 1-Mb region of NBS-LRR genes (MELO3C004235 to MELO3C004331) re-annotated in chromosome 5 of melon genome (improved sequence 5,190,204–6,256,576 from CM 3.5: 5,190,204–6,256,576) [73]. This variations should be considered in future to improve marker development. Further, our reported R gene-related markers especially for PM disease associated genes will be suitable for design genotyping by sequencing (GBS), linkage map construction, and development of PM disease-associated QTLs/genes. In addition, our suggested makers from this study provide support for marker-assisted breeding in melon and highlight areas for further study to understand the functional significance of each accession.
Supporting Information
S1 Fig. Response of melon accessions to powdery mildew (Podosphaera xanthii).
Shown are the leaf disc tests after inculcation (A) and measured infected areas (B) of young leaves with representative infection severity.
https://doi.org/10.1371/journal.pone.0157524.s001
(TIF)
S2 Fig. Distribution of the number of InDels identified in melon accessions.
https://doi.org/10.1371/journal.pone.0157524.s002
(TIF)
S3 Fig.
Chromosomal distribution of detected (A) SNP and (B) InDel variations in melon pseudo-chromosomes.
https://doi.org/10.1371/journal.pone.0157524.s003
(TIF)
S4 Fig.
GO classification of detected high- and moderate-effect SNPs in (A) SCNU1154, (B) Edisto47, (C) MR-1, and (D) PMR5 accessions.
https://doi.org/10.1371/journal.pone.0157524.s004
(TIF)
S5 Fig.
GO classification of detected high- and moderate-effect InDels in (A) SCNU1154, (B) Edisto47, (C) MR-1, and (D) PMR5 accessions.
https://doi.org/10.1371/journal.pone.0157524.s005
(TIF)
S6 Fig.
Comparative analysis of (A) SNP and (B) InDel variants in MLO genes.
https://doi.org/10.1371/journal.pone.0157524.s006
(TIF)
S1 Table. Frequency of transitions and transversions in each accession.
https://doi.org/10.1371/journal.pone.0157524.s007
(XLSX)
S2 Table. The distribution of InDel lengths in each accession.
https://doi.org/10.1371/journal.pone.0157524.s008
(XLSX)
S3 Table. Distribution of SNP/InDel effects by type and region across melon accessions.
https://doi.org/10.1371/journal.pone.0157524.s009
(XLSX)
S4 Table. Chromosomal distribution of detected structure variations.
https://doi.org/10.1371/journal.pone.0157524.s010
(XLSX)
S8 Table. SNP and InDel variations in MLO genes.
https://doi.org/10.1371/journal.pone.0157524.s014
(XLSX)
Acknowledgments
This research was supported by Golden Seed Project (Center for Horticultural Seed Development, Ministry of Agriculture, Food and Rural Affairs (MAFRA), Ministry of Oceans and Fisheries (MOF), Rural Development Administration (RDA) and Korea Forest Service (KFS): 213003-04-4-SB110. Export Promotion Technology Development Program (Grant No. 312065-05-4-HD030), Ministry of Agriculture, Food and Rural Affairs, Republic of Korea.
Author Contributions
Conceived and designed the experiments: ISN JIP. Analyzed the data: SN. Wrote the paper: SN. Provided support for experimental validation and data analysis: HTK SKT KV.
References
- 1. Jeffrey C. A review of the Cucurbitaceae. Botanical Journal of the Linnean Society. 1980;81(3):233–47.
- 2.
Pitrat M. Melon. In: Prohens J, Nuez F, editors. Vegetables I. Handbook of Plant Breeding. 1: Springer New York; 2008. p. 283–315.
- 3. Sebastian P, Schaefer H, Telford IRH, Renner SS. Cucumber (Cucumis sativus) and melon (C. melo) have numerous wild relatives in Asia and Australia, and the sister species of melon is from Australia. Proceedings of the National Academy of Sciences. 2010;107(32):14269–73.
- 4. Pech J-C, Bouzayen M, Latché A. Climacteric fruit ripening: ethylene-dependent and independent regulation of ripening pathways in melon fruit. Plant Science. 2008;175(1):114–20.
- 5. Martin A, Troadec C, Boualem A, Rajab M, Fernandez R, Morin H, et al. A transposon-induced epigenetic change leads to sex determination in melon. Nature. 2009;461(7267):1135–8. pmid:19847267
- 6. Zhang B, Tolstikov V, Turnbull C, Hicks LM, Fiehn O. Divergent metabolome and proteome suggest functional independence of dual phloem transport systems in cucurbits. Proceedings of the National Academy of Sciences. 2010;107(30):13532–7.
- 7. González VM, Garcia-Mas J, Arús P, Puigdomènech P. Generation of a BAC-based physical map of the melon genome. BMC genomics. 2010;11(1):339.
- 8. Diaz A, Fergany M, Formisano G, Ziarsolo P, Blanca J, Fei Z, et al. A consensus linkage map for molecular markers and quantitative trait loci associated with economically important traits in melon (Cucumis melo L.). BMC plant biology. 2011;11(1):111.
- 9. Clepet C, Joobeur T, Zheng Y, Jublot D, Huang M, Truniger V, et al. Analysis of expressed sequence tags generated from full-length enriched cDNA libraries of melon. BMC genomics. 2011;12(1):252.
- 10. Dahmani-Mardas F, Troadec C, Boualem A, Leveˆque S, Alsadon AA, Aldoss AA, et al. Engineering melon plants with improved fruit shelf life using the TILLING approach. PLoS One. 2010;5(12):e15776. pmid:21209891
- 11. González VM, Benjak A, Hénaff EM, Mir G, Casacuberta JM, Garcia-Mas J, et al. Sequencing of 6.7 Mb of the melon genome using a BAC pooling strategy. BMC plant biology. 2010;10(1):246.
- 12. Mascarell-Creus A, Cañizares J, Vilarrasa-Blasi J, Mora-García S, Blanca J, Gonzalez-Ibeas D, et al. An oligo-based microarray offers novel transcriptomic approaches for the analysis of pathogen resistance and fruit quality traits in melon (Cucumis melo L.). BMC genomics. 2009;10(1):467.
- 13.
Kong Q, Yuan J, Niu P, Xie J, Jiang W, Huang Y, et al. Screening suitable reference genes for normalization in reverse transcription quantitative real-time PCR analysis in melon. 2014.
- 14. McGrath MT. Fungicide resistance in cucurbit powdery mildew: experiences and challenges. Plant disease. 2001;85(3):236–45.
- 15. Sitterly WR. Powdery mildews of cucurbits. The powdery mildews. 1978:359–79.
- 16. Lebeda A, Křístková E, Sedláková B, McCreight JD, Coffey MD. Cucurbit powdery mildews: methodology for objective determination and denomination of races. European Journal of Plant Pathology. 2015:1–12.
- 17.
Jahn M, Munger H, McCreight J, Bélanger R, Bushnell W, Dik A, et al. Breeding cucurbit crops for powdery mildew resistance. The powdery mildews: a comprehensive treatise. 2002:239–48.
- 18.
Lebeda A, Sedláková B. Identification and survey of cucurbit powdery mildew races in Czech populations. Cucurbitaceae 2006, Asheville, North Carolina, USA, 17–21 September 2006. 2006:444–52.
- 19.
Dogimont C. 2011 Gene List for Melon.
- 20. Silberstein L, Kovalski I, Huang R, Anagnostou K, Jahn MMK, Perl-Treves R. Molecular variation in melon (Cucumis melo L.) as revealed by RFLP and RAPD markers. Scientia Horticulturae. 1999;79(1):101–11.
- 21. Garcia-Mas J, Oliver M, Gomez-Paniagua H, De Vicente M. Comparing AFLP, RAPD and RFLP markers for measuring genetic diversity in melon. Theoretical and Applied Genetics. 2000;101(5–6):860–4.
- 22. Esteras C, Formisano G, Roig C, Díaz A, Blanca J, Garcia-Mas J, et al. SNP genotyping in melons: genetic variation, population structure, and linkage disequilibrium. Theoretical and Applied Genetics. 2013;126(5):1285–303. pmid:23381808
- 23. Monforte A, Garcia‐Mas J, Arús P. Genetic variability in melon based on microsatellite variation. Plant breeding. 2003;122(2):153–7.
- 24. Xu X, Bai G. Whole-genome resequencing: changing the paradigms of SNP detection, molecular mapping and gene discovery. Molecular Breeding. 2015;35(1):1–11.
- 25. Lai J, Li R, Xu X, Jin W, Xu M, Zhao H, et al. Genome-wide patterns of genetic variation among elite maize inbred lines. Nature genetics. 2010;42(11):1027–30. pmid:20972441
- 26. Garcia-Mas J, Benjak A, Sanseverino W, Bourgeois M, Mir G, González VM, et al. The genome of melon (Cucumis melo L.). Proceedings of the National Academy of Sciences. 2012;109(29):11872–7.
- 27. Guo S, Zhang J, Sun H, Salse J, Lucas WJ, Zhang H, et al. The draft genome of watermelon (Citrullus lanatus) and resequencing of 20 diverse accessions. Nature genetics. 2013;45(1):51–8. pmid:23179023
- 28. Huang S, Li R, Zhang Z, Li L, Gu X, Fan W, et al. The genome of the cucumber, Cucumis sativus L. Nature genetics. 2009;41(12):1275–81. pmid:19881527
- 29. Kim HT, Park JI, Nou IS. Identification of fungal races that cause powdery mildew in melon (Cucumis melo L.) and selection of resistant commercial melon cultivars against the identified races in Korea. Journal of Plant Biotechnology. 2016;43(1):58–65.
- 30. MORISHITA M, SUGIYAMA K, SAITO T, SAKATA Y. Powdery mildew resistance in cucumber. Japan Agricultural Research Quarterly: JARQ. 2003;37(1):7–14.
- 31. Fukino N, Kunihisa M, Matsumoto S. Characterization of Recombinant Inbred Lines Derived from Crosses in Melon (Cucumis melo L.),'PMAR No. 5'*'Harukei No. 3'. Breeding science. 2004;54(2):141–5.
- 32. Salari M, Panjehkeh N, Nasirpoor Z, Abkhoo J. Reaction of melon (Cucumis melo L.) cultivars to soil-borne plant pathogenic fungi in Iran. African Journal of Biotechnology. 2012;11(87):15324–9.
- 33. Li H, Durbin R. Fast and accurate short read alignment with Burrows–Wheeler transform. Bioinformatics. 2009;25(14):1754–60. pmid:19451168
- 34. McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A, et al. The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome research. 2010;20(9):1297–303. pmid:20644199
- 35. Ruffalo M, LaFramboise T, Koyutürk M. Comparative analysis of algorithms for next-generation sequencing read alignment. Bioinformatics. 2011;27(20):2790–6. pmid:21856737
- 36. Cingolani P, Platts A, Wang LL, Coon M, Nguyen T, Wang L, et al. A program for annotating and predicting the effects of single nucleotide polymorphisms, SnpEff: SNPs in the genome of Drosophila melanogaster strain w1118; iso-2; iso-3. Fly. 2012;6(2):80–92. pmid:22728672
- 37. Conesa A, Götz S, García-Gómez JM, Terol J, Talón M, Robles M. Blast2GO: a universal tool for annotation, visualization and analysis in functional genomics research. Bioinformatics. 2005;21(18):3674–6. pmid:16081474
- 38. Kanehisa M, Goto S. KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res. 2000;28(1):27–30. Epub 1999/12/11. gkd027 [pii]. pmid:10592173; PubMed Central PMCID: PMC102409.
- 39. Finn RD, Bateman A, Clements J, Coggill P, Eberhardt RY, Eddy SR, et al. Pfam: the protein families database. Nucleic Acids Research. 2014;42(D1):D222–D30.
- 40. Sanseverino W, Hermoso A, D’Alessandro R, Vlasova A, Andolfo G, Frusciante L, et al. PRGdb 2.0: towards a community-based database model for the analysis of R-genes in plants. Nucleic Acids Research. 2012.
- 41. Voorrips RE. MapChart: Software for the Graphical Presentation of Linkage Maps and QTLs. Journal of Heredity. 2002;93(1):77–8. pmid:12011185
- 42. Chen K, Wallis JW, McLellan MD, Larson DE, Kalicki JM, Pohl CS, et al. BreakDancer: an algorithm for high-resolution mapping of genomic structural variation. Nat Meth. 2009;6(9):677–81. http://www.nature.com/nmeth/journal/v6/n9/suppinfo/nmeth.1363_S1.html.
- 43. Krzywinski M, Schein J, Birol I, Connors J, Gascoyne R, Horsman D, et al. Circos: an information aesthetic for comparative genomics. Genome research. 2009;19(9):1639–45. pmid:19541911
- 44. Nunes EWLP, Preston HAP, DoVale JC, Aragão FASd, Antônio RP, Nunes GHdS. New sources of resistance to Myrothecium roridum and Podosphaeria xanthii in yellow melon. Crop Breeding and Applied Biotechnology. 2015;15(3):181–6.
- 45. Causse M, Desplat N, Pascual L, Le Paslier M-C, Sauvage C, Bauchet G, et al. Whole genome resequencing in tomato reveals variation associated with introgression and breeding events. BMC genomics. 2013;14(1):791.
- 46. Saxena RK, Edwards D, Varshney RK. Structural variations in plant genomes. Brief Funct Genomics. 2014;13(4):296–307. Epub 2014/06/08. [pii]. pmid:24907366; PubMed Central PMCID: PMC4110416.
- 47. Pitrat M, Lecoq H. Inheritance of zucchini yellow mosaic virus resistance in Cucumis melo L. Euphytica. 1984;33(1):57–61.
- 48.
Quiot-Douine L, Lecoq H, Quiot J, Pitrat M, Labonne G, editors. Evidence for a biological and serological variability in a potyvirus infecting cucurbits: the papaya ringspot virus (PRSV). EUCARPIA meeting on cucurbit genetics and breeding; 1988.
- 49. Pitrat M. Linkage Groups in Cucumis melo L. Journal of Heredity. 1991;82(5).
- 50.
Webb R. Inheritance of resistance to watermelon mosaic virus 1 in Cucumis melo L.[Cantaloupes]. HortScience (USA). 1979.
- 51. Dogimont C, Bussemakers A, Martin J, Slama S, Lecoq H, Pitrat M. Two complementary recessive genes conferring resistance to Cucurbit Aphid Borne Yellows Luteovirus in an Indian melon line (Cucumis melo L.). Euphytica. 1997;96(3):391–5.
- 52.
Esteva J, Nuez F, Cuartero J. Resistance to Yellowing Disease in Wild Relatives of Muskrnelon. 1988.
- 53. López-Sesé A, Gomez-Guillamon M. Resistance to Cucurbit yellowing stunting disorder virus (CYSDV) in Cucumis melo L. HortScience. 2000;35(1):110–3.
- 54.
McCreight JD. Potential sources of resistance to lettuce infectious yellows virus in melon. Cucurbit Genetic Cooperative. 1991;Report 14::51–2.
- 55. McCreight JD, Liu H-Y, Turini TA. Genetic resistance to Cucurbit leaf crumple virus in melon. HortScience. 2008;43(1):122–6.
- 56. Gumpf D, Kishaba A, Bohn G. Identification, seed transmission, and host range pathogenicity of a California isolate of melon necrotic spot virus. Phytopathology. 1979.
- 57. Sugiyama M, Ohara T, Sakata Y. Inheritance of resistance to Cucumber green mottle mosaic virus in Cucumis melo L. ‘Chang Bougi’. Journal of the Japanese Society for Horticultural Science. 2007;76::316–8.
- 58. Brotman Y, Kovalski I, Dogimont C, Pitrat M, Portnoy V, Katzir N, et al. Molecular markers linked to papaya ring spot virus resistance and Fusarium race 2 resistance in melon. Theoretical and Applied Genetics. 2005;110(2):337–45. pmid:15551034
- 59. Baudracco-Arnas S, Pitrat M. A genetic map of melon (Cucumis melo L.) with RFLP, RAPD, isozyme, disease resistance and morphological markers. Theor Appl Genet. 1996;93::57–64. pmid:24162199
- 60. Joobeur T, King JJ, Nolin SJ, Thomas CE, Dean RA. The fusarium wilt resistance locus Fom-2 of melon contains a single resistance gene with complex features. Plant Journal. 2004;39::283–97. pmid:15255859
- 61. Jagger I, Whitaker T, Porter D. Inheritance in Cucumis melo of resistance to powdery mildew (Erysiphe cichoracearum). Phytopathology. 1938;28:761.
- 62. Bohn GW, Whitaker TW. Genetics of resistance to powdery mildew race 2 in muskmelon. Phytopathology. 1964;54(5):587–91.
- 63. Harwood RR, Markarian D. A genetic survey of resistance to powdery mildew in muskmelon. Journal of Heredity. 1968;59(3):213–7.
- 64. HARWOOD RR, MARKARIAN D. The inheritance of resistance to powdery mildew in the cantaloupe variety Seminole. Journal of Heredity. 1968;59(2):126–30.
- 65. KENIGHSBUCH D, Cohen Y. Independent inheritance of resistance to race 1 and race 2 of Sphaerotheca fuliginea in muskmelon. Plant disease. 1989;73(3):206–8.
- 66. Liu L, Chen Y, Su Z, Zhang H, Zhu W. A sequence-amplified characterized region marker for a single, dominant gene in melon PI 134198 that confers resistance to a unique race of Podosphaera xanthii in China. HortScience. 2010;45(9):1407–10.
- 67. Epinat C, Pitrat M, Bertrand F. Genetic analysis of resistance of five melon lines to powdery mildews. Euphytica. 1992;65(2):135–44.
- 68. Iovieno P, Andolfo G, Schiavulli A, Catalano D, Ricciardi L, Frusciante L, et al. Structure, evolution and functional inference on the Mildew Locus O (MLO) gene family in three cultivated Cucurbitaceae spp. BMC genomics. 2015;16(1):1112. Epub 2015/12/31. [pii]. pmid:26715041; PubMed Central PMCID: PMC4696115.
- 69. Gonzalez V, Aventin N, Centeno E, Puigdomenech P. High presence/absence gene variability in defense-related gene clusters of Cucumis melo. BMC genomics. 2013;14(1):782.
- 70. Leida C, Moser C, Esteras C, Sulpice R, Lunn J, de Langen F, et al. Variability of candidate genes, genetic structure and association with sugar accumulation and climacteric behavior in a broad germplasm collection of melon (Cucumis melo L.). BMC Genetics. 2015;16(1):28.
- 71. Jewel H, Hoy-Taek K, Jong-In P, Nasar Uddin A, Arif Hasan Khan R, Hee-Jeong J, et al. Expression Profiling of MLO Family Genes under Podosphaera xanthii Infection and Exogenous Application of Phytohormones in Cucumis melo L. 생명과학회지. 2016;26(4):419–30.
- 72. Ning X, Wang X, Gao X, Zhang Z, Zhang L, Yan W, et al. Inheritances and location of powdery mildew resistance gene in melon Edisto47. Euphytica. 2014;195(3):345–53.
- 73. González VM, Aventín N, Centeno E, Puigdomènech P. Interspecific and intraspecific gene variability in a 1-Mb region containing the highest density of NBS-LRR genes found in the melon genome. BMC genomics. 2014;15(1):1.