Development of Molecular Marker Linked with Bacterial Fruit Blotch Resistance in Melon (Cucumis melo L.)

Bacterial fruit blotch (BFB) causes losses in melon marketable yield. However, until now, there has been no information about the genetic loci responsible for resistance to the disease or their pattern of inheritance. We determined the inheritance pattern of BFB resistance from a segregating population of 491 F2 individuals raised by crossing BFB-resistant (PI 353814) and susceptible (PI 614596) parental accessions. All F1 plants were resistant to Acidovorax citrulli strain KACC18782, and F2 plants segregated with a 3:1 ratio for resistant and susceptible phenotypes, respectively, in a seedling bioassay experiment, indicating that BFB resistance is controlled by a monogenic dominant gene. In an investigation of 57 putative disease-resistance related genes across the melon genome, only the MELO3C022157 gene (encoding TIR-NBS-LRR domain), showing polymorphism between resistant and susceptible parents, revealed as a good candidate for further investigation. Cloning, sequencing and quantitative RT-PCR expression of the polymorphic gene MELO3C022157 located on chromosome 9 revealed multiple insertion/deletions (InDels) and single nucleotide polymorphisms (SNPs), of which the SNP A2035T in the second exon of the gene caused loss of the LRR domain and truncated protein in the susceptible accession. The InDel marker MB157-2, based on the large (504 bp) insertion in the first intron of the susceptible accession, was able to distinguish resistant and susceptible accessions among 491 F2 and 22 landraces/inbred accessions with 98.17% and 100% detection accuracy, respectively. This novel PCR-based, co-dominant InDel marker represents a practical tool for marker-assisted breeding aimed at developing BFB-resistant melon accessions.


Introduction
Melon (Cucumis melo L., 2n = 2× = 24) belonging to the Cucurbitaceae family produce delicious, sweet, fruits rich in vitamins, minerals and health-promoting antioxidants [1]. They are an economically important agricultural crop cultivated and consumed all around the world, with an average production of more than 29 million tons per year throughout the past decade [2]. Melons are predominantly cultivated in temperate and tropical countries with all the major producing regions in the world (China, Turkey, Iran, India, Kazakhstan, USA, Egypt, Spain, Guatemala, Italy, etc.) (http://www.fao.org/ faostat/en/#rankings/countries_by_commodity). Production is greatly hampered by several pathogens creating a barrier to attaining higher yields of this important fruit crop.

Bacterial Culture and Inoculum Preparation
A. citrulli strain KACC18782 was obtained from the Korean Agricultural Culture Collection (KACC), South Korea. Bacteria were grown in King's B (KB) media [3] supplemented with 100 µg mL −1 ampicillin for 36-48 h at 28 • C until formation of bacterial colonies. For all inoculations, a bacterial suspension was prepared by inundating culture plates with 5 mL of sterile, double-distilled water and gently scraping the surface of the KB media using a sterile, L-shaped rubber spreader to an optical density of 1.0 at 600 nm, as measured using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA). Bacterial suspension was diluted to a final concentration of~1 × 10 6 colony forming units (mL −1 ).

Inoculation, Sample Preparation and Assessment of BFB Resistance
Plants with 3-5 true leaves (4-5 weeks old) were inoculated by spraying with a suspension of A. citrulli KACC18782 using a hand-pump spray bottle until runoff in a greenhouse with a temperature range of 20-25 • C. Inoculated plants were covered with a plastic cover to maintain high relative humidity (96%). Plants were re-inoculated 3 days after the first inoculation to ensure that no plants had escaped inoculation and to eliminate false positives. Samples BFB-resistant and susceptible plants were collected from the fourth and fifth true leaves stage of the plants at the time points of 12 h, 1 day, 3 days and 6 days after inoculation and from control plants at the same time points. The samples were immediately frozen in liquid nitrogen (N 2 ). The disease severity of three individual leaves from each inoculated plant was scored 12 days after inoculation using a scale of 1 to 6 (Supplementary Figure S1) for ≥10%, 11-20%, 21-50%, 51-75%, 76-90% and >90% diseased/infected area per leaf, respectively [18]. Disease ratings were scored as per the methods described by Robin et al. [36] and Hassan et al. [37] with little modification. The percentage infected area (PIA) was measured as the ratio of infected area to total leaf area, multiplied by 100. Individuals with a PIA of ≤20 and >21 were considered as resistant and susceptible, respectively [18,37]. Supplementary Table S2 and the resulting ratio was tested for goodness-of-fit using χ 2 analysis.
Genes 2020, 11, 220 6 of 18 Moreover, for the quantitative RT-PCR expression analysis of polymorphic (MELO3C022157) gene-specific primers were designed F: CGAATTTGAGTGCTGTTCCA and R: CGAAGTTCTACTGTTGGGCG on the fourth exon. Primers (Table 1), one pair for shorter genes and multiple pairs for longer genes, were designed to cover the entire length of the chromosomes using the Primer3Plus (http://www.bioinformatics.nl/cgi-bin/primer3plus/primer3plus.cgi) web tool. Primer sequences were checked for potential self-dimer and hairpin formation using an oligo calculator (http://biotools.nubic.northwestern.edu/OligoCalc.html). PCR was carried out in a thermo-cycler with initial denaturation at 95 • C for 5 min, followed by 35 cycles of denaturation at 95 • C for 30 s, annealing at 58 • C for 30 s and elongation at 72 • C for 45 s and final elongation at 72 • C for 7 min. Electrophoresis of amplified PCR products was performed in 1.2% agarose gel stained with HiQ blue mango (20,000×) (bioD, Gwangmyaong, Korea) and visualized using ultraviolet light in a ENDURO™ GDS gel documentation system (New York, NY, USA).

Cloning and Sequencing of the Polymorphic Gene
The polymorphic gene MELO3C022157 was PCR-amplified from resistant (PI 353814) and susceptible (PI 614596) accession using Phusion ® High-Fidelity DNA Polymerase (New England Biolabs, EVRY Cedex, France). Amplified DNA fragments were purified using a Promega DNA Purification kit (Promega, Madison, WI, USA). Cloning was performed using a TOPO-TA cloning kit (Invitrogen, Carlsbad, CA, USA) following the manufacturer's instructions. The universal primers M13F and M13RpUC were used to sequence cloned amplicons using an ABI3730XL sequencer (Macrogen Co., Seoul, Korea). Each forward and reverse sequence of resistant and susceptible melon accession was repeated three times to remove all uncertainties. Gene sequences of the resistant and susceptible accessions were aligned using CLUSTALW software (https://www.genome.jp/tools-bin/clustalw) to detect sequence variation.

Total RNA Extraction and cDNA Synthesis
Control and infected melon leaves were crushed to a powder in liquid nitrogen (N 2 ), and 100 mg of each sample was subjected to total RNA extraction using an RNeasy Mini kit (Qiagen, Valencia, CA, USA) following the manufacturer's instructions. First-strand cDNA was synthesized from total RNA with a SuperScript III First-Strand Synthesis System kit (Invitrogen, Gaithersburg, MD, USA).

Quantitative RT-PCR (qRT-PCR) Expression of Polymorphic Gene MELO3C022157
The expression patterns of the polymorphic-gene were analyzed by qRT-PCR in a LightCycler ® instrument (Roche, Mannheim, Germany) following the manufacturer's instructions. The reactions were performed in a 10-µL volume consisting of 5 µL of 2x qPCRBIO SyGreen Mix Lo-ROX (PCR Biosystems, London, UK); 5 pmol of primers and cDNA templates diluted to the appropriate concentrations. The PCR conditions were as follows: 5 min at 95 • C, followed by 3-step amplifications at 95 • C for 15 s, 56 • C for 15 s and 72 • C for 20 s, for 45 cycles. The mean expression level of relevant genes was calculated by the 2 −∆∆ Ct method [38], where Actin was used as an internal control [39].

Statistical Analysis
To determine significant changes in gene expression levels among the different time points for each treatment, a chi-square (χ 2 ) test for goodness-of-fit was performed to determine deviations of observed data from expected segregation ratios using Minitab18 statistical software (State College, PA, USA). A value of p < 0.05 was considered statistically significant. Turkey's pairwise comparison test was conducted for mean separation.

Dominant Inheritance of BFB Resistance in Melon
To determine the inheritance pattern of BFB resistance in melon, we crossed a BFB-resistant (PI 353814) and a BFB-susceptible (PI 614596) parental accession for making F 1 . The resistant parental accessions PI 353814 had less than 20% BFB-infected leaf area, while more than 70% of the leaves of susceptible parental accessions PI 614596 were infected by A. citrulli (KACC18782) (Figure 1). Symptoms of the leaves of F 1 hybrids resembled those of the resistant parent, indicating the dominant nature of the inheritance of this trait in melons. We next evaluated the inheritance of BFB resistance in the F 2 population consisting of 491 plants.

Statistical Analysis
To determine significant changes in gene expression levels among the different time points for each treatment, a chi-square (χ 2 ) test for goodness-of-fit was performed to determine deviations of observed data from expected segregation ratios using Minitab18 statistical software (State College, PA, USA). A value of p < 0.05 was considered statistically significant. Turkey's pairwise comparison test was conducted for mean separation.

Dominant Inheritance of BFB Resistance in Melon
To determine the inheritance pattern of BFB resistance in melon, we crossed a BFB-resistant (PI 353814) and a BFB-susceptible (PI 614596) parental accession for making F1. The resistant parental accessions PI 353814 had less than 20% BFB-infected leaf area, while more than 70% of the leaves of susceptible parental accessions PI 614596 were infected by A. citrulli (KACC18782) (Figure 1). Symptoms of the leaves of F1 hybrids resembled those of the resistant parent, indicating the dominant nature of the inheritance of this trait in melons. We next evaluated the inheritance of BFB resistance in the F2 population consisting of 491 plants. Bioassay results indicated that 360 F2 plants were resistant, and 131 were susceptible. A chisquare (χ 2 ) test revealed that BFB resistance segregated with a 3:1 ratio (resistant: susceptible), indicating monogenic dominant control of the trait (Table 2 and Supplementary Figures S2 and S3).

Microsynteny Analysis of 57 R-Genes of Melon Compared with Watermelon and Cucumber
We performed comparative analysis to identify the homologous R-genes among melon, compared with watermelon and cucumber. Some R-genes from melons share homologous relationships with those of watermelon and cucumber. However, only four chromosome genes of melons (chr.01, chr.03, chr.05 and chr.09) share homologous relationships with those of watermelon, while remaining chromosome genes are not homologous. Moreover, genes of melon on chr.06, chr.07 and chr.12 lack homologues in watermelon, whereas other chromosome genes are homologous (Figure 2).

Microsynteny Analysis of 57 R-Genes of Melon Compared with Watermelon and Cucumber
We performed comparative analysis to identify the homologous R-genes among melon, compared with watermelon and cucumber. Some R-genes from melons share homologous relationships with those of watermelon and cucumber. However, only four chromosome genes of melons (chr.01, chr.03, chr.05 and chr.09) share homologous relationships with those of watermelon, while remaining chromosome genes are not homologous. Moreover, genes of melon on chr.06, chr.07 and chr.12 lack homologues in watermelon, whereas other chromosome genes are homologous ( Figure 2).

Identification of Length Polymorphisms in Putative R-genes
It is essential to identify BFB-resistant genotypes to develop BFB-resistant melon cultivars. Since no definitive R-gene is known to be associated with BFB resistance in melon, we identified polymorphisms in putative R-genes with NBS, LRR, TIR and CC domains throughout the genome based on a functional annotation of sequences in the Cucurbit Genomics database (http://cucurbitgenomics.org/search/genome/3).
We selected 57 NBS-encoding genes covering all chromosomes of melon, with a maximum 17 and 9 genes on chromosomes 9 and 1, respectively (Table 1). Among these 57 genes, 44 were NBS-LRR (NL) type, while 17 and 2 were TIR-NBS-LRR (TNL) and TIR-NBS (TN) types, respectively. Only one gene each was selected from the NBS and CC-NBS-LRR (CNL) types. We designed gene-specific

Identification of Length Polymorphisms in Putative R-genes
It is essential to identify BFB-resistant genotypes to develop BFB-resistant melon cultivars. Since no definitive R-gene is known to be associated with BFB resistance in melon, we identified polymorphisms in putative R-genes with NBS, LRR, TIR and CC domains throughout the genome based on a functional annotation of sequences in the Cucurbit Genomics database (http://cucurbitgenomics.org/search/genome/3).
We selected 57 NBS-encoding genes covering all chromosomes of melon, with a maximum 17 and 9 genes on chromosomes 9 and 1, respectively (Table 1). Among these 57 genes, 44 were NBS-LRR (NL) type, while 17 and 2 were TIR-NBS-LRR (TNL) and TIR-NBS (TN) types, respectively. Only one gene each was selected from the NBS and CC-NBS-LRR (CNL) types. We designed gene-specific primers (one set for shorter genes and multiple sets for longer genes) covering the entire length of these genes for PCR amplification and detection of length polymorphisms between resistant and susceptible accessions. Among these genes, conspicuous length polymorphism was detected only for the disease resistance-related TNL-type gene MELO3C022157 (Supplementary Figure S4).

Cloning and Sequencing of the Polymorphic Gene
We divided the polymorphic gene MELO3C022157 into six consecutive fragments amplified by six sets of primers (Table 3) to narrow down the specific polymorphic region. The polymorphic region was identified in the first intron of the gene ( Figure 3; Figure 4b). We cloned and sequenced the six MELO3C022157 gene fragments from both resistant and susceptible accessions, revealing insertion of a 504 bp fragment in the first intron of the susceptible parent (Figure 4c). primers (one set for shorter genes and multiple sets for longer genes) covering the entire length of these genes for PCR amplification and detection of length polymorphisms between resistant and susceptible accessions. Among these genes, conspicuous length polymorphism was detected only for the disease resistance-related TNL-type gene MELO3C022157 (Supplementary Figure S4).

Cloning and Sequencing of the Polymorphic Gene
We divided the polymorphic gene MELO3C022157 into six consecutive fragments amplified by six sets of primers (Table 3) to narrow down the specific polymorphic region. The polymorphic region was identified in the first intron of the gene ( Figure 3; Figure 4b). We cloned and sequenced the six MELO3C022157 gene fragments from both resistant and susceptible accessions, revealing insertion of a 504 bp fragment in the first intron of the susceptible parent (Figure 4c).  Table 3.   Table 3.
Genes 2020, 11, 220 13 of 18 primers (one set for shorter genes and multiple sets for longer genes) covering the entire length of these genes for PCR amplification and detection of length polymorphisms between resistant and susceptible accessions. Among these genes, conspicuous length polymorphism was detected only for the disease resistance-related TNL-type gene MELO3C022157 (Supplementary Figure S4).

Cloning and Sequencing of the Polymorphic Gene
We divided the polymorphic gene MELO3C022157 into six consecutive fragments amplified by six sets of primers (Table 3) to narrow down the specific polymorphic region. The polymorphic region was identified in the first intron of the gene ( Figure 3; Figure 4b). We cloned and sequenced the six MELO3C022157 gene fragments from both resistant and susceptible accessions, revealing insertion of a 504 bp fragment in the first intron of the susceptible parent (Figure 4c).  Table 3.   In addition, we detected a 5 bp deletion in the 1st. intron of the resistant accession; a 4 bp deletion in the 2nd. exon of the susceptible accession; three insertions: 1 bp in the 2nd. exon and 3 and 1 bp in the 3rd. exon of the susceptible accession and several single nucleotide polymorphisms (SNPs: "A/G, A/T, A/G and A/T), all "A" in the resistant parent at positions 1355, 1526, 2454 and 2893 in the 2nd., 2nd., 3rd. and 4th. exons, respectively, and "G","T", "G" and "T" in the susceptible parent at positions 1864, 2035, 2959 and 3398 in the 2nd. , 2nd., 3rd. and 4th. exons, respectively, and "G","T" and "G" in the susceptible parent at positions 1864, 2035 and 2959 in the 2nd. , 2nd. and 3rd. exons, respectively (Figure 4a and Supplementary Figure S6b). The SNP "A/T" ("A" and "T" in resistant and susceptible parents at positions 1526 and 2035, respectively) generates a premature stop codon in the susceptible parent that produces a truncated protein (Figure 4d, Supplementary Figure S6b,c). In silico domain analysis of the translated protein sequences of the cloned gene of resistant and susceptible accessions using InterProScan (https://www.ebi.ac.uk/interpro/search/sequence-search) revealed that this SNP causes loss of the LRR domain in the susceptible accessions PI 614596, whereas this is present in the resistant accessions PI 353814 (Figure 4e).

qRT-PCR Expression Analysis of Polymorphic Gene MELO3C022157
The expression of the polymorphic gene via qRT-PCR gradually increased in the leaf tissue of the resistant (PI 353814), as compared to the susceptible (PI 614596) parent at different time points. The gene was induced within 12 h of infection by A. citrulli in both the resistant and susceptible parents. In general, the transcript level in the resistant parent was higher at 12 h, 1 d, 3 d and 6 d after inoculation followed by a decrease in expression ( Figure 5) and, after 6d, expression level was very low in the susceptible parent. The gene expression levels revealed a 5-fold peak 6d after inoculation in the resistant parent, whereas transcript levels were lowest in the susceptible parent at this time point.
Genes 2020, 11, 220 11 of 18 In addition, we detected a 5 bp deletion in the 1st. intron of the resistant accession; a 4 bp deletion in the 2nd. exon of the susceptible accession; three insertions: 1 bp in the 2nd. exon and 3 and 1 bp in the 3rd. exon of the susceptible accession and several single nucleotide polymorphisms (SNPs: "A/G, A/T, A/G and A/T), all "A" in the resistant parent at positions 1355, 1526, 2454 and 2893 in the 2nd., 2nd., 3rd. and 4th. exons, respectively, and "G","T", "G" and "T" in the susceptible parent at positions 1864, 2035, 2959 and 3398 in the 2nd., 2nd., 3rd. and 4th. exons, respectively, and "G","T" and "G" in the susceptible parent at positions 1864, 2035 and 2959 in the 2nd., 2nd. and 3rd. exons, respectively (Figure 4a and Supplementary Figure S6b). The SNP "A/T" ("A" and "T" in resistant and susceptible parents at positions 1526 and 2035, respectively) generates a premature stop codon in the susceptible parent that produces a truncated protein (Figure 4d, Supplementary Figure S6b,c). In silico domain analysis of the translated protein sequences of the cloned gene of resistant and susceptible accessions using InterProScan (https://www.ebi.ac.uk/interpro/search/sequence-search) revealed that this SNP causes loss of the LRR domain in the susceptible accessions PI 614596, whereas this is present in the resistant accessions PI 353814 (Figure 4e).

qRT-PCR Expression Analysis of Polymorphic Gene MELO3C022157
The expression of the polymorphic gene via qRT-PCR gradually increased in the leaf tissue of the resistant (PI 353814), as compared to the susceptible (PI 614596) parent at different time points. The gene was induced within 12 h of infection by A. citrulli in both the resistant and susceptible parents. In general, the transcript level in the resistant parent was higher at 12 h, 1 d, 3 d and 6 d after inoculation followed by a decrease in expression ( Figure 5) and, after 6d, expression level was very low in the susceptible parent. The gene expression levels revealed a 5-fold peak 6d after inoculation in the resistant parent, whereas transcript levels were lowest in the susceptible parent at this time point.

Development and Validation of InDel Marker Linked to BFB Resistance in Melon
Among the polymorphisms, we targeted the long insertion/deletion (InDel) mutation between resistant and susceptible accessions and designed a set of primers (MB157-2-F/R) around this mutation (Table 3) by producing a 287 bp amplicon for the resistant accessions, a 791 bp amplicon for the susceptible accessions and both bands for the heterozygous F1 hybrid (Figure 4b). To validate the efficiency of the marker, we genotyped 491 F2 individuals generated from the resistant and

Development and Validation of InDel Marker Linked to BFB Resistance in Melon
Among the polymorphisms, we targeted the long insertion/deletion (InDel) mutation between resistant and susceptible accessions and designed a set of primers (MB157-2-F/R) around this mutation (Table 3) by producing a 287 bp amplicon for the resistant accessions, a 791 bp amplicon for the susceptible accessions and both bands for the heterozygous F 1 hybrid (Figure 4b). To validate the efficiency of the marker, we genotyped 491 F 2 individuals generated from the resistant and susceptible accessions. The genotypes of 482 individuals matched those from the bioassay results, indicating a detection accuracy of about 98.17% ( Figure 6 and Supplementary Table S2). We then used the marker to genotype 22 landraces/inbred melon accessions collected from various sources ( Supplementary Table S1), of which we detected two accessions, PI 140471 and PI 420145, as resistant ( Figure S5b). A bioassay test of these 22 landraces/inbred accessions indicated that these two accessions were indeed resistant to BFB ( Figure S5a). These results suggest that the InDel MB157-2 marker can be effectively used for detecting resistant and susceptible melon genotypes using a PCR-based assay.
susceptible accessions. The genotypes of 482 individuals matched those from the bioassay results, indicating a detection accuracy of about 98.17% ( Figure 6 and Supplementary Table S2). We then used the marker to genotype 22 landraces/inbred melon accessions collected from various sources ( Supplementary Table S1), of which we detected two accessions, PI 140471 and PI 420145, as resistant ( Figures S5b). A bioassay test of these 22 landraces/inbred accessions indicated that these two accessions were indeed resistant to BFB ( Figure S5a). These results suggest that the InDel MB157-2 marker can be effectively used for detecting resistant and susceptible melon genotypes using a PCRbased assay.

Discussion
Developing BFB-resistant melon accessions is essential, since chemical control measures and other cultural practices are only partially effective, increase the cost of production, are not sustainable and are hazardous to the environment [40]. Even though BFB causes extensive damage, there has been no genetic information about any causal loci or genes controlling the resistants or their pattern of inheritance. Here, we determined the inheritance of BFB disease in melon and developed an InDel marker for distinguishing resistant and susceptible melon genotypes through a PCR assay.
Field evaluation after inoculation with A. citrulli strain KACC18782 indicated that seedlings of all F1 plants from the cross PI 614596 × PI 353814 were resistant, demonstrating the completely dominant nature of the trait in melon (Figure 1). Similar evaluations of 491 F2 individuals showed a Mendelian segregation phenotypic ratio of 3:1 (resistant:susceptible) and the genotypic ratio of 1:2:1 (homozygous (resistant):heterozygous (resistant):homozygous (susceptible)), indicating monogenic control of resistance to the disease (Table 2 and Supplementary Figures S2 and S3).

Discussion
Developing BFB-resistant melon accessions is essential, since chemical control measures and other cultural practices are only partially effective, increase the cost of production, are not sustainable and are hazardous to the environment [40]. Even though BFB causes extensive damage, there has been no genetic information about any causal loci or genes controlling the resistants or their pattern of inheritance. Here, we determined the inheritance of BFB disease in melon and developed an InDel marker for distinguishing resistant and susceptible melon genotypes through a PCR assay.
Field evaluation after inoculation with A. citrulli strain KACC18782 indicated that seedlings of all F 1 plants from the cross PI 614596 × PI 353814 were resistant, demonstrating the completely dominant nature of the trait in melon (Figure 1). Similar evaluations of 491 F 2 individuals showed a Mendelian segregation phenotypic ratio of 3:1 (resistant:susceptible) and the genotypic ratio of 1:2:1 (homozygous (resistant):heterozygous (resistant):homozygous (susceptible)), indicating monogenic control of resistance to the disease (Table 2 and Supplementary Figures S2 and S3).
We then focused on developing molecular markers to detect resistant vs. susceptible genotypes by a simple PCR assay. However, since no QTL or functional gene has been reported for BFB resistance in melon, we examined putative R-genes genome-wide.
We investigated length polymorphism in 57 putative disease resistance genes encoding proteins containing NBS, LRR, TIR and CC domains and covering all 12 chromosomes of melon, with particular emphasis on chromosomes and regions harboring disease and insect resistance-related loci. Conspicuous length polymorphism between resistant and susceptible melon accessions was only observed for the TIR-NBS-LRR gene MELO3C022157 (Table 1 and Supplementary Figure S4). Besides, the presence of homologues of this R-gene in watermelon and cucumber ( Figure 3) indicates that this gene is likely to play a similar role in melon [35,37]. After cloning and sequencing of the entire polymorphic gene (resistant and susceptible parent) revealed a long insertion (504 bp) in the first intron in the susceptible parent and several other small InDels and SNPs between the resistant and susceptible parents (Figure 4c and Supplementary Figure S6b). Of these, the SNP A 2035 T in the second exon caused loss of the LRR domain, resulting in a truncated protein in the susceptible accession (Figure 4e and Supplementary Figure S6c). So, the polymorphic gene belongs to the LRR domain; qRT-PCR transcription level is higher at 6d after inoculation in resistants, whilst loss of LRR and transcription level is little at 6d in the susceptible (Figures 4e and 5), indicating that 6d after inoculation in the susceptible parent fails to develop a hypersensitive response, i.e., causing BFB disease, suggesting that the LRR domain may be responsible for the susceptible reaction of the mutant accession. LRR domains are involved in resistance against a yellow strain of cucumber mosaic virus (CMV[Y]) in Arabidopsis [57,58], which strongly support our results. LRR domains also implicate LepR3 loci mediate resistance against blackleg in Brassica napus [59] and against Fusarium wilt (Fom2) in melon [57,58]. Since LRR domains provide specificity in pathogen recognition [48,60], the NBS-LRR class of plant innate immune receptors uses its LRR domain to accomplish many other roles [61].
In developing a molecular marker, we targeted the 504 bp InDel region in the first intron of MELO3C022157 and designed primer pair MB157-2-F/R, which identified resistant and susceptible accessions among 491 F 2 individuals with 98.17% detection accuracy and those representing 22 landraces/inbred accessions with 100% detection accuracy (Table 3, Supplementary Tables S1 and S2 and Figure 6 and Supplementary Figure S5b). Introns play diverse roles in processes such as transcription coupling [62], gene expression regulation, formation of noncoding RNA [63,64] and exon shuffling and alternative splicing [65]. In addition, intronic polymorphism is a valuable source for developing genetic markers with high interspecies transferability [66,67], since introns are under much lower selective pressure than other functional elements such as exons [68]. Several markers have been developed using intronic polymorphism for a range of traits in different crops, including cabbage, melon, maize and sunflower [37,66,69,70]. Notably, a pair of InDel markers located in the same first intron of MELO3C022157 are reportedly linked to gummy stem blight resistance in melon [37].

Conclusions
We determined that BFB resistance in melon is controlled by a single dominant gene, allowing the development of the PCR-based co-dominant InDel marker MB157-2 for identifying resistant and susceptible accessions. The co-dominant InDel marker MB157-2 developed in this study will have practical implications for marker-assisted breeding for improving BFB resistance in melons. This is the first report, to our knowledge, of a molecular marker linked to BFB resistance in melons. Work is underway to map the resistant loci using partial genome sequence-based approaches.
Supplementary Materials: The following are available online at http://www.mdpi.com/2073-4425/11/2/220/s1; Table S1: Details of 22 melon accessions used and their disease response, as determined by bioassay and PCR-based assay using polymorphic InDel marker MB157-2. Table S2: Comparison of Acidovorax citrulli bioassay results and InDel marker BM157-2 bands for the putative R-gene MELO3C022157 in a population of 491 F2 melon individuals raised from resistant and susceptible parental accessions PI 353814 and PI 614596, respectively. "+", "и" and "-" indicate resistant, heterozygous and susceptible bands for the InDel marker BM157-2. Figure S1: Disease scores used to assess the severity of bacterial fruit blotch (BFB) caused by A. citrulli at 12 days after inoculation. Scores range from 1 to 6, with 1 representing leaves with no damage and 6 representing maximum damage. All leaves were detached just before photographs were taken. Figure S2: Disease symptoms on resistant (R) and susceptible (S) melon parental accessions PI 353814 and PI 614596, respectively, their F1 hybrid and 491 individuals of the F2 population at 12 days after inoculation with A. citrulli. All leaves were detached just before photographs were taken. Figure S3: Frequency distribution of BFB scores in melon P 1 , P 2 , F 1 and F 2 populations. Figure S4: Detection of length polymorphism in 57 putative disease resistance genes containing NBS, LRR, CC and TIR domains by PCR-based assay. Details of the genes and corresponding primer specifications are presented in Table 1. Genes 40-2 and 57 were not amplified. Gene 41 (red underline) showed conspicuous length polymorphism. R resistant parent PI 353814, S susceptible parent PI 614596 and F1 hybrid (PI 614596 × PI 353814). Figure S5: (a) Bacterial fruit blotch (BFB) disease symptoms on 22 melon accessions at 12 days after inoculation with A. citrulli. Details of these 22 melon accessions are shown in Supplementary Table S1. All leaves were detached just before photographs were taken. (b) Validation of resistance status to BFB using InDel marker MB157-2 in the 22 melon accessions. R resistant (PI 353814) and S susceptible (PI 614596) used as control. Figure