Genome-wide surveying of cucumber (Cucumis sativus L.) lines for idetifying loci of ZYMV resistance

doi:10.21203/rs.3.rs-1739512/v1

Control of virus diseases such as zuchini yellow mosaic virus (ZYMV) of cucumber causing major yield and quality losses is impossible by using any chemical pesticide except for using elimination of vectors of viruses and resistant cultivar. The objectives of this study were to investigate structure of genetic resistance to ZYMV of cucumber for marker-assisted selection (MAS). Forty-eigth resistant and 48 susceptible lines were selected from a germplasm of 600 cucumber lines. Bulk DNAs of resistant and susceptible lines were screened to find candidate marker(s) by using ISSR, RAPD and a cleaved amplified polymorphism (CAP) primers, which targeted over 5000 loci of cucumber. Candidate markers later were applied to 96 lines including 48 resistance and susceptible lines to find associating markers. A total of 54 candidate marker loci were produced, which were clustered into two groups based on Dice’, average distance, simple matching and correlation similarity matrices. In the association analyses by using TASSEL software, associating markers were detected. These markers were then subjected to backward regression analysis by SAS program. The results explained %78 of total variation for ZYMV. The remaining 22% of the variation was possibly due to presence of additional gene(s) and environmental effect that caused lower repeatability.

Cucumber

GWAS

ZYMV

MAS

virus resistance

biotic stress

Virus diseases cause significant crop losses in cucumber (Cucumis sativus L.). Zucchini yellow mosaic virüs (ZYMV) is an aphid-borne potyvirus and one of the most common pathogens causing major losses worldwide in the members of the Cucurbitaceae family (Yılmaz and Davis 1985; Brunt et al. 1996; Gal-On 2007; Kamberoğlu et al. 2016). This virus has various strains affecting all cucurbits. The measures against these diseases are limited (Lecoq et al. 2004). Genetic resistance of cultivars to the pathogens, chemical pesticides against vectors carrying viruses and avoidance of entering insect vectors into production areas by nets are some of the options (Yılmaz et al. 1994; Coutts et al. 2011). In addition, weak strain applications through cross-protection may provide partial protection in few countries (Yılmaz et al. 1994). A recent approach, transgenic cucumber through gene silencing, may also provide solution to eliminate effect of this pathogen, but the acceptability of transgenic crops is likely to create concern for future utilization (Leibman et al. 2011).

Breeding for ZYMV resistance is the most desirable approach. Inheritance of this trait has been previously characterized (Providenti 1996; Cardoso et al. 2010). These findings suggested that ZYMV resistance in cucumber is inherited recessively. Thus, introduction of resistance allele into elite lines requires generations of backcrossing and inoculations of progenies with pathogens, extending the time required for developing resistant plants. This time can be shortened using marker-assisted selection (MAS). Several DNA markers linked to the recessive ZYMV resistance gene (zym) have been previously reported (Park et al. 2004). These markers were not polymorphic between some ZYMV resistant and susceptible lines as also detected in our previous analysis of the germplasm (not published). Amano et al. (2013) developed six candidate genes flanked by two SSR markers within zymA192-18 locus to a < 50- kb genomic region, which needed confirmation over different gene pools of cucumber lines available in the collections of public and private institutions. Recently, Monthizuki et al. (2016) have found that infection with ZYMV of cucumber facilitated entry of cucumber mosaic virus into xylem. This may indicate that ZYMV interacts with other internal and external factors. Seemingly, different loci for ZYMV resistance are likely to exist. Therefore, additional investigation linked to the zym loci covering highly variable germplasm are needed for marker development. The objective of this study was to investigate structure of genetic resistance to ZYMV of cucumber for MAS.

Plant materials and DNA isolation

More than 600 accessions selfed 4-6 times were previously screened for ZYMV resistance by classical method verified by ELISA. Forty-eight resistant and susceptible lines were selected from this population for further analyses. Seeds of each accession were sown in plastic multi-pots (4 × 4 × 4 cm) filled with peat:perlite (2:1) for germination and bulk of fresh true leaves from each accession were harvested into freezer bags and kept at -20^oC until DNA isolation. Total DNA was extracted from 30 mg tissue by using a modified CTAB DNA extraction procedure (Gulsen et al. 2010). DNA pellets were diluted with 300 ul of TE (10 mM Tris, 0.1 mM EDTA, pH 7.4). For PCR reactions, 10 ng/ ul DNA templates were made using double-distilled water and checked on 2% agarose gels.

DNA analyses and marker production

Equal amounts of DNAs of five resistant and susceptiple lines were mixed to prepare two bulks, which were later used to screened for molecular marker polymorphism. For marker production, 1160 RAPD and 26 ISSR, and one CAP primers were used to evaluate resistant and susceptible bulks. Those primers producing marker polymorphism between the bulks were applied to all 96 samples (48 resistant and 48 susceptible). A total of 26 ISSR primers [(GA)₈YG, VHVG(TA)₇, (CAC)₆, DBDA(CA)₇, (GT)₈YA, BDB(CA)₇, (GAA)₆, (TAA)₈, (GACA)₄, (CA)₆AC, (AG)₇YC, (CAC)₃GC, (CA)₈, (GT)₆GG, (AGC)₆G, (CT)₈TG, (AG)₈T, VHVG(TG)₇, BAB(CA)₇C, (HVG(CA)₇T, (TCC)₅RY, UBC807-891, (AG)₈YC, (CT)₉RC, BDB(CA)₇C and (CAC)₆]were used for screening bulks. A CAP marker developed by Amano et al. 2013) was also used for analysis and its sequence is as follows: dCAPS-G99A; forward primer 5’-ACGCAAAAGCCTCTCCGCTGTATTT-3’, reverse primer 5’-GCTCTCCAATCCAGCAACAT-3’. This produced 338 bp PCR fragment, then digested with DraI enzyme producing 313 and 25 bp fragments.

Each of 15 ul PCR components consisted of 0.66 mM of each of primers, 200 uM of each dNTPs, 1.5 ul of 10× PCR buffer, 2–2.5 mM of MgCl₂, ddH₂O, one unit of Taq polymerase and 20 ng of template DNA. PCR products of four marker systems were separated on 2-3% agarose gel at 110 V for 3–6 h and visualized under UV light.

Data analyses

Firts, we performed cluster analyses of markers in order to see clustering pattern of 54 markers produced from bulks of resistant and susceptible lines. This may provide possible linkages among the markers since they tend to be inherited more frequently. Each band was scored as present (1) or absent (0) and data were analyzed with the Numerical Taxonomy Multivariate Analysis System (NTSYS-pc version 2.1) software package (Rohlf, 2000). A similarity matrix was constructed based on simple matching, correlation, average distance and Dice’s coefficiencies (Dice, 1945). These similarity matrices were used to construct a dendrogram using the unweighted pair group method arithmetic average (UPGMA) to determine genetic relationships among the germplasm studied.

Population structure was analyzed using a model-based approach, Bayesian method by STRUCTURE software version 2.3.2 (Pritchard, 2000). Output of this analysis was used as covariate to avoid spurious linkages. Model-based cluster analysis was used to test whether K = 1–10, where K is the number of sub-populations. There is informal pointer to detect the best K, in which values of log Pr (X/K) reach more or less plateaus after major decrease. Admixture and independent allele frequencies were used. For each population (K), 100000 iteration and 100000 burn-in period options were used. For each number of K from 1 to 10, five independent calculations were performed, and likelihood values obtained from these calculations were averaged for each K. Finally, a graph was drawn by using Microsoft Office Excel for visual observation. Membership coefficients produced by the STRUCTURE program were used to infer whether the accessions are admixed by number of subpopulations or pure. Association analyses between marker loci and phenotypic values were performed using the general linear model (GLM) functions in TASSEL stand-alone version 3.1 (Bradbury et al. 2007). Marker loci with more than 15% occurrence for a given marker loci were used in the analyses. To calculate associations between studied characters and polymorphic bands by TASSEL, three separate data files (trait, molecular marker data and population structure) were prepared. Initially, filtered structure file was created by selecting one subpopulation. Then, three files were joined by using JOIN option of TASSEL software. Two different p-values, F test- and permutation-based, were calculated between each marker and character, which provided significantly associating markers with the trait interest at alpha 0.05. Type I error was controlled by applying the 1000 permutation test in TASSEL stand-alone version 3.1 to control error associated with multiple comparisons and generate strong p-values for association analysis of polymorphic sites. Finally, backward regression was used to develop regression model to predict phenotype based on genotype by using SAS software. Backward regression starts with all candidate variables, testing the deletion of each variable using a chosen model comparison criterion, deleting the variable (if any) that improves the model the most by being deleted, and repeating this process until no further improvement is possible.

Production of polymorphic molecular markers

A total of 1187 primers including 1160 RAPD, 26 ISSR and one CAP primers were used in this study. These primers were first applied to bulk DNAs of susceptible and resistant individuals, then to all 96 samples. This amplified more than 5000 thousand of loci of Cucumis sativus. Among 1160 RAPD primers, 106 were applied to the 96 samples, images of 40 of which were scored for the data analyses. Of the 26 ISSR primers produced 13 polymorphic fragments among 96 resistant and susceptible lines. In addition, one CAP marker digested with DraI enzyme produced a codominant polymorphism among the lines.

Assessment of previously developed CAP Marker

Amano et al. (2013) reported a CAP marker (dCAPS-G99A) cosegregating with resistance. This marker was analyzed for the lines/cultivars in this study (Fig. 1) High level of association was observed between this CAP marker and resistance. On the other hand, although the samples 54, 71, 72 and 74 were susceptible, they had resistance allele. Similarly, among the resistant samples as detected based on classical screening, lines 80 and 250 were heterozygous, and 272, 91, 125, 147, 148, 379, 341, 345, 188, 203 and 214 had susceptible allele. The CAP marker developed by Amano et al. (2013) was also assessed for some other known cultivars and lines in this study (Figure not shown). The known resistant (Silyon RZ, TMG-1, Sardes) and susceptible (Maraton and 8 other lines) were tested with this CAP marker. Silyon and TMG-1 had resistant allele, but Sardes had susceptible allele of the CAP marker. All susceptible samples had susceptible allele.

Investigation of grouping pattern among the molecular markers

The cluster analysis was performed for 54 markers to detect associations among the loci. Three out of four different matrices (correlation, simple matching, average distance, and Dice’ similarity) clustered the 54 markers into two subclusters (Fig. 2).

Detection of molecular markers associating with ZYMV and regression models

Membership coefficiencies to five subpopulations produced by STRUCTURE program were used as covariance in TASSEL software. Out of 54 markers, only loosely correlating 38 as detected by NTSYS software were used for substructuring Bayesian analysis, which resulted in five subpopulations among the 96 samples. For estimating candidate markers for ZYMV resistance of cucumber, association analysis was performed by using three data files (phenotype, genotype, and substructure) in TASSEL. Based of F-statistics and permutation (1000), candidate markers were detected (Table 1). The CAP marker alone explained 42% of the total variation for ZYMV among the 96 samples, followed by OPM07600 (27%) ve OPD07920 (26%).

All markers revealed 80% of the variation for ZYMV resistance in cucumber. Of all markers, those significantly associating markers based on both F- and permutation-test revealed 78 and 75% of the total variation before and after BR (P < 0.0001) (Table 2). After BR, only five markers remained in the regression model (Table 3).

Table 2. Statistical parameters after analysis of associating markers based on both F- and permutation-test before BR.

Parameter Standard

Variable Estimate Error Type II SS F Value Pr > F

Intercept 0.31224 0.22164 0.14088 1.98 0.1647

CAP 0.38710 0.09458 1.18907 16.75 0.0001

OPD07920 -0.34371 0.08486 1.16448 16.40 0.0002

OPE16770 0.02635 0.11580 0.00368 0.05 0.8209

OPAL111000 -0.02165 0.09394 0.00377 0.05 0.8186

OPM07600 -0.02018 0.12310 0.00191 0.03 0.8704

OPN041000 -0.04665 0.12645 0.00966 0.14 0.7137

OPBE121100 -0.08672 0.11323 0.04164 0.59 0.4472

OPBA131800 -0.09503 0.10517 0.05796 0.82 0.3703

OPBA16670 -0.11463 0.11143 0.07511 1.06 0.3083

OPU012000 -0.28118 0.11940 0.39369 5.55 0.0223

OPAN011100 0.15363 0.10160 0.16231 2.29 0.1364

GT8YA1000 0.07055 0.10776 0.03043 0.43 0.5155

AG8YC500 0.21421 0.07410 0.59324 8.36 0.0056

OPP12550 0.12550 0.07626 0.19222 2.71 0.1058

Table 3. Statistical parameters after analysis of associating markers based on both F- and permutation-test after BR.

Parameter Standard

Variable Estimate Error Type II SS F Value Pr > F

Intercept 0.37720 0.19271 0.26356 3.83 0.0548

CAP 0.41901 0.08051 1.86344 27.09 <.0001

OPD07920 -0.35448 0.07564 1.51098 21.97 <.0001

OPBE121100 -0.15010 0.07062 0.31079 4.52 0.0375

OPU012000 -0.37963 0.09419 1.11751 16.25 0.0002

AG8YC500 0.21707 0.07007 0.66016 9.60 0.0029

Although previous studies reporting monogenic control of ZYMV resistance of cucumber (Providenti 1996; Cardoso et al. 2010), our recent findings have indicated that the markers developed for resistance locus failed to distinguish resistant and susceptible lines as discussed above. We screened large germplasm collection to elucidate genetic structure of resistance to ZYMV among cucumbers including lines and popular cultivars worldwide. The data of the classical screening in our previous study and molecular screening in this study were combined to estimate genetic structure of ZYMV resistance in cucumber. The classical screening was previously performed under greenhouse conditions in Antalya and verified by ELISA tests. This study was repeated twice and each repeat had 10 replications. The molecular data were produced in current study. A total of 1187 primers (1160 RAPD, 26 ISSR and one CAP primers) targeting approximately more than 5.000 loci were investigated in this study. Of these primers, 110 primers were applied to all 96 samples and produced 54 polymorphic markers. These efforts aimed whole genome survey for ZYMV resistance in cucumber.

Based on 54 markers polymorphic between the bulks and among 48 susceptible and 48 resistant cucumbers, the cluster analysis was performed to detect associations among the marker loci. Three out of four different matrices (correlation, simple matching, average distance and Dice’ similarity) clustered the 54 markers into two subclusters (Fig. 2). This pointed polygenic control of ZYMV resistance in cucumbers since closely linked markers tended to be clustered in proximity.

Association analyses of these markers with ZYMV resistance in cucumber resulted in significant findings on genetic structure of ZYMV resistance in cucumber. The CAP marker developed by Amano et al. (2013) revealed only 42% of the variation for ZYMV resistance. The markers developed in our study increased this level up to 75%. The literature review discussed above indicated single locus for ZYMV resistance. Our study indicated, though, possible other loci based on both cluster analyses of 54 markers and regression analyses (Fig. 4). The reason behind this contradiction is probably due to size of germplasm used in this study. We surveyed larger gene pool of cucumber by using genome-wide association study, while previous studies employed single population derived from two parents only. The larger gene pool in our germplasm including more than cucumber 600 lines may likely present larger number of loci related to ZYMV resistance.

Other important point to discuss is level of efficiency of classical screening study. Due to many factors such as age of plant, temperetaure, pathogen concentration and other uncontrollable conditions, results may be significantly affected. Classical screening was repeated twice and for some samples, screening was repeated more. Despite these efforts, some escape is likely. Therefore, we strongly believe that increasing the efficiency of classical screening will also increase the level variation explained by molecular markers developed in this study.

Although single locus for ZYMV resistance in cucumber was reported, along with major locus/loci, modifying gene(s) partially affecting disease progress in the tissues of cucumber is/are possible. The ZYMV markers reported by Park et al. (2004) did not distinguish the samples used in this study for ZYMV resistance, while it was successful in their study. The marker reported by Amano et al. (2013), on the other hand, revealed 42% of the variation for ZYMV resistance. We also tested the known resistant (Silyon RZ, TMG-1, Sardes) and susceptible (Maraton and 8 others known) lines with this CAP marker. Silyon and TMG-1 had resistant allele, but Sardes had susceptible allele of the CAP marker. All susceptible samples had susceptible allele. These findings support our conclusion that ZYMV is polygenic in this study.

The markers developed in this study may allow successful discrimination of resistant and susceptible cucumbers for markers assisted selection in breeding programs. Here we may suggest the following regression model to estimate ZYMV resistance in cucumber breeding programs.

Phenotype_ZYMV = 0.37720 + CAP (0.41901) - OPD07920 (-0.35448) - OPBE121100 (-0.15010) - OPU012000 (-0.37963) + AG8YC500 (0.21707).

Acknowledgments

This study was financially supported by the section of Scientific Research Unit of Erciyes University (FBA-12-3883). YMV isolate, known for virulence was kindly provided by the Virology Laboratory of the University of Akdeniz. Few of the cucumber lines for CAP marker analysis were kindly provided by Manier Seed Company, Adana, Turkey.

Conflict of interest The Authors declare that they have no confict of interest whatsoever in this study. This article does not contain any studies with animals performed by any of the authors.

Amano M, Mochizuki A, Kawagoe Y, Iwahori K, Niwa K, Svoboda J, Maeda T, Imura Y (2013) High-resolution mapping of zym, a recessive gene for Zucchini yellow mosaic virus resistance in cucumber. Teor Applied Genet. 126: 2983-2993. DOI: 10.1007/s00122-013-2187-5.

Bradbury PJ, Zhang Z, Kroon DE, Casstevens TM, Ramdos Y, Bucklar ES (2007) TASSEL: software for association mapping of complex traits in diverse samples. Bioinformatics. 23: 2633-2635. DOI: 10.1093/bioinformatics/btm308.

Brunt AA, Crapte K, Dallywitz M, Gibs A, Watson L (1996) Viruses of plants description and lists from the wide database. University Press, Cambridge, U.K.

Cardoso AII, Pavan MA, Sakate RK (2010) Inheritance of cucumber tolerance to zucchini yellow mosaic virus. J. Plant Pathology. 92: 245-248. DOI: 10.1007/BF00223299.

Coutts BA, Kehoe MA, Jones RAC (2011) Minimising losses caused by Zucchini yellow mosaic virus in vegetable cucurbit crops in tropical, sub-tropical and Mediterranean environments through cultural methods and host resistance. Virus Res. 159: 141-160. DOI: 10.1016/j.virusres.2011.04.015.

Dice LR (1945). Measures of the amount of ecologic association between species. Ecology. 26: 297-302.

Gal-On A (2007) Zucchini yellow mosaic virus: insect transmission and pathogenicity - the tails of two proteins. Mol Plant Pathology. 8: 139-150. DOI: 10.1111/j.1364-3703.2007.00381.x.

Gulsen O, Kaymak S, Ozongun S, Uzun A (2010) Genetic analysis of Turkish apple germplasm using peroxidase gene-based markers. Scientia Horticulturae. 125: 368–373. DOI: 10.1016/j.scienta.2010.04.023

Kamberoglu MA, Caliskan AF, Desbiez C (2016) Current status of some cucurbit viruses in Cukurova region (Adana and Mersin provinces) of Turkey and molecular characterization of zucchini yellow mosaic virus isolates. Romanian Biotech L. 21: 11709-11719.

Lecoq H, Moury B, Desbiez C, Palloix A, Pitrat M (2004) Durable virus resistance in plants through conventional approaches: a challenge. Virus Research. 100: 31-39. https://doi.org/10.1016/j.virusres.2003.12.012.

Leibman D, Wolf D, Saharan V, Zelcer A, Arazi T, Yoel S, Gaba V, Gal-On A (2011) High level of transgenic viral small RNA is associated with broad potyvirus resistance in cucurbits. Mol Plant-Microbe Int. 24: 1220-1238. DOI: 10.1094/MPMI-05-11-0128.

Mochizuki T, Nobuhara S, Nishimura M, Ryang Bo-Song Naoe M, Matsumoto T, Kosaka Y, Ohki S (2016) The entry of cucumber mosaic virus into cucumber xylem is facilitated by co-infection with zucchini yellow mosaic virüs. Archives of Virol. 161: 2683-2692. DOI: 10.1007/s00705-016-2970-0.

Park Y, Katzir N, Brotman Y, King J, Bertrand F, Havey M (2004) Comparative mapping of ZYMV resistances in cucumber (Cucumis sativus L.) and melon (Cucumis melo L.). Theor Applied Genet. 109:707-712. DOI: 10.1007/s00122-004-1684-y

Pritchard JK, Stephens M, Rosenberg NA, Donnely P (2000) Association mapping in structured populations. Am J Hum Genet. 67: 170-181. DOI: 10.1086/302959.

Providenti R (1991) Inheritance of resistance to the Florida strain of Zucchini yellow mosaic virus in watermelon. HortScience 26, 407–408.

Rohlf FJ (2000) NTSYS-pc, Numerical Taxonomy and Multivariate Analysis System. Version 2.11. New York, Exeter, Setauket.

Yılmaz MA, Davis RF (1985) Identification of viruses infecting vegetable crops along the Mediterranean sea cost in Turkey. J. Turk Phytopathol. 14:1-8.

Yılmaz MA, Abak K, Lecoq H, Baloglu S, Sari N (1994) Control of zucchini yellow mosaic virus in cucurbits by ZYMV-WK starin. Proc. 9^th Cong. Mediterranean Phytopathol. Union, Kusadası, Turkey, pp:353-356.

Table 1 not available with this version.

No competing interests reported.

Genome-wide surveying of cucumber (Cucumis sativus L.) lines for idetifying loci of ZYMV resistance

Status:

Version 1

Abstract

Figures

Introduction

Materials And Methods

Results

Discussion

Declarations

References

Tables

Competing Interests

Status:

Version 1