Begomovirus Tomato Leaf Curl New Delhi Virus Is Seedborne but Not Seed Transmitted in Melon

Transmission of plant viruses through seed can be one of the major factors contributing to long-distance dispersal through global trade of seeds and can have important ecological consequences for virus dissemination (Johansen et al. 1994; Jones 2021; Maule and Wang 1996). Seed transmission of plant viruses at epidemiologically relevant levels is not the norm. Nevertheless, about 18% of plant viruses are transmitted through seeds in one or more hosts (Johansen et al. 1994; Sastry 2013). For cultivated plants that are grown from seed, seed transmission can facilitate virus perpetuation between growing seasons and the emergence of viruses in areas where they were not previously present.

Transmission by seeds can be achieved through external contamination of the seed coat, with emerging seedlings becoming infected through microlesions during germination, as reported for tobamoviruses (genus Tobamovirus, family Virgaviridae) (Dombrovsky and Smith 2017; Hull 2014). Alternatively, the virus could invade the seed embryo through mechanisms largely unknown and move to offspring tissues during germination, as shown for members of the Potyviridae family (Simmons and Munkvold 2014). In the latter case, seed transmission will take place if the virus invades plant reproductive organs at the correct times to access embryo tissues (Hull 2014; Maule and Wang 1996).

Begomoviruses (genus Begomovirus, family Geminiviridae) (Zerbini et al. 2017) and, among them, isolates of the species Tomato leaf curl New Delhi virus, cause significant yield losses in economically important crops (Rojas et al. 2018). These viruses are transmitted in nature in a circulative and persistent mode by members of the whitefly Bemisia tabaci (Genn.) cryptic species (Hemiptera: Aleyrodidae) (De Barro et al. 2011; de Moya et al. 2019). Thus, the worldwide expansion of this invasive whitefly has been considered a key factor in the emergence of begomoviruses (Gilbertson et al. 2015; Rojas et al. 2018). However, in recent years, several reports have raised the possibility of transmission through seeds for some members of this genus (Kim et al. 2015; Kothandaraman et al. 2016; Manivannan et al. 2019; Suruthi et al. 2018). No transmission through seeds was demonstrated in other cases (Pérez-Padilla et al. 2020; Sisodia and Mahatma 2020; Tabein et al. 2021).

Tomato leaf curl New Delhi virus (ToLCNDV) is a bipartite begomovirus with two similar-sized, single-stranded circular DNA genome components (named DNA-A and DNA-B) of approximately 2.8 kb in size, which threatens economically important vegetable and fiber crops. ToLCNDV was initially reported causing losses in tomato but, subsequently, many reports also included species of the cucurbit family as major hosts for this virus. In addition to tomato and cucurbits, this virus causes damage to other relevant crops and infects several weed species (Moriones et al. 2017). Although epidemics of ToLCNDV were originally limited to Asia, a recent westward spread of the virus has occurred, affecting areas in the Middle East and the Mediterranean Basin in quite a short time period, with severe disease outbreaks reported in cucurbit crops of Spain, Italy, Tunisia, Morocco, and Algeria (Kheireddine et al. 2019; Moriones et al. 2017; Sifres et al. 2018). Sequence analysis suggested that isolates of a strain of ToLCNDV (“Spain” strain, ToLCNDV-ES), whose diversification has involved recombination (Fortes et al. 2016), are spreading in the Western Mediterranean area. The recent long-distance and rapid westward spread of ToLCNDV very much mirrors the movement of isolates of the begomovirus tomato yellow leaf curl virus (TYLCV) (a member of the species Tomato yellow leaf curl virus) across the Mediterranean and other geographic areas in the 1990s, which was associated with the B. tabaci emergence or live planting material movement (Lefeuvre et al. 2010; Polston et al. 1999; Rojas et al. 2018). Recent reports have raised the possibility that seed transmission of ToLCNDV in cucurbits may occur (Sangeetha et al. 2018; Kil et al. 2020).

Melon (Cucumis melo L.) is one of the most important cucurbit crops in the world and it is associated with an intense and economically important seed trade. Because infections of ToLCNDV have been reported in this plant host (López et al. 2015), this study was undertaken to address the potential for ToLCNDV infection or contamination of melon seeds (seedborne), and the subsequent transmission to plants produced from the seed (seed transmission). In either scenario, the implications to both disease management and practices employed in melon seed production by the seed industry would require reconsideration. Studies were conducted with three different melon cultivars (all susceptible to ToLCNDV). ToLCNDV presence in floral tissues and mature seeds of infected plants, and the possible seed-to-progeny plant transmission, were investigated under controlled, insect-proof conditions.

Materials and Methods

Plant materials, virus isolate, and seed collection

Three commercial hybrid F₁ melon cultivars were used in this study: one Galia type, cultivar Brimos (named as BR), and two Yellow types, cultivars Mayor (MY) and Nesta (Nst). Seeds collected from ToLCNDV-free melon plants were kindly provided by Semillas Fitó, Spain. The infectious clone of the isolate ES-Alm-661-Sq-13 of the “Spain” (ES) strain of ToLCNDV (ToLCNDV-Spain[ES-Alm-661-Sq-13]; DNA-A and DNA-B GenBank accession numbers KF749223 and KF749226, respectively) previously described by Fortes et al. (2016) was used for experimental inoculation.

Mature seeds were collected from fruit harvested from melon plants of the cultivars BR, MY, and Nst experimentally inoculated with ToLCNDV at the two- to three-leaf stage with the infectious ToLCNDV clone by stem puncture using Agrobacterium tumefaciens-mediated inoculation (agroinoculation) (Monci et al. 2005). Agroinoculation was carried out in a growth chamber (26 and 22°C, day and night, respectively, and 70% relative humidity, with a 16-h photoperiod of photosynthetically active radiation at 250 mol s⁻¹ m⁻²) and the inoculated plants were then transferred to an insect-proof glasshouse with temperature control (approximately 16-h day length at 22 to 27°C during the day and 17 to 20°C at night, with supplemental light when needed) until fruit harvest. Mature melon seeds were removed from completely ripe fruit (45 to 50 days postanthesis) by thorough washing with water (nontreated seeds = NT) as recommended for obtaining optimal germinability in cucurbits (George 2009; Nerson 2002). Part of the seeds was then additionally surface treated following two different procedures: (i) washing thoroughly by stirring (1,500 rpm) for 10 min using a magnetic stirrer in distilled water followed by two additional washes (2 min each) in distilled water (water removed and water-treated seeds = WT) and (ii) surface disinfection by immersion and stirring (1,500 rpm) for 10 min in 50% commercial bleach (about 1.5% sodium hypochlorite) followed by two thorough washes (2 min each) in distilled water (water removed and surface-disinfected seeds = WD). Seeds were then dried at 25°C in a heater with air recirculation for 24 h to complete drying and stored at 4°C until used. The seed germination was evaluated by sowing 20 melon seeds per treatment (NT, WD, and WT), with similar germination efficiency observed in each case. The seeds collected from ToLCNDV-free melon plants were used as a negative experimental control.

Detection of ToLCNDV presence in melon plants, flowers, and seeds

ToLCNDV detection in plants was done by hybridization of tissue blots from the youngest leaf petiole cross-sections (squash-blot hybridization) using a probe specific to ToLCNDV and the protocol according to Fortes et al. (2016). Viral detection was also conducted on total DNA extracted from the youngest leaf of tested plants, seed, or flower samples either by conventional PCR amplification or by real-time quantitative PCR (qPCR) amplification. Information about the primers used in the latter cases is summarized in Table 1. When needed, bulk samples from leaves (0.2 g per individual plant) or seeds were used for total DNA extraction and subsequent PCR or qPCR analysis. The seeds were ground (1 min at 30 rpm) with a steel bead using the homogenizer Retsch Mixer Mill MM-400 (GmbH, Haan, Germany). The ground seed samples were mixed thoroughly with 500 µl of phosphate-buffered saline (PBS, pH 7.4) and centrifuged at 20,000 × g for 3 min at 4°C. Then, 100-µl aliquots of the supernatant from each seed sample were collected for subsequent DNA extraction. Individual or bulk samples of leaves or flowers were homogenized in a Stomacher 80 Biomaster/Filter Bags (Seward Ltd., Worthing, U.K.) in 5 ml of PBS supplemented with Tween-20 (0.02%); homogenized samples were held at 4°C during processing until DNA extraction from a volume of 100 µl per sample. Total DNA was extracted from seed, leaf, or flower samples following a modified cetyltrimethylammonium bromide method (Permingeat et al. 1998). For conventional PCR, the primer pairs ToLCNDV-A-F/ToLCNDV-A-R or ToLCNDV-F/ToLCNDV-R, specific to DNA-A of ToLCNDV (size of amplification products: 1,260 and 120 bp, respectively) were used. Amplification of the melon cytochrome c oxidase (COX) gene (COX-F/COX-R; amplified fragment: 100 bp) was used as control. Conventional PCR amplification was carried out using a T100 Thermal Cycler (Bio-Rad, Hercules, CA, U.S.A.) using the cycle parameters 3 min at 95°C; followed by 30 cycles of 30 s at 95°C, 45 s at 56°C (57°C for primer pair ToLCNDV-A-F/ToLCNDV-A-R), and 30 s (1 min for primer pair ToLCNDV-A-F/ToLCNDV-A-R) at 72°C; with a final incubation at 72°C for 10 min. The reactions were carried out in 20 µl of final volume of PCR mix containing 2 µl of 10× NH4 Reaction Buffer (Meridian Bioscience Inc., Cincinnati, OH, U.S.A.), 1.5 mM MgCl₂, 0.2 mM each dNTP, 0.2 µM each primer, 1.5 U of Biotaq DNA Polymerase (Meridian Bioscience Inc.), and 100 ng of extracted total DNA as template. Viral DNA accumulation was determined on total DNA extracts by qPCR amplification with the primer pair and TaqMan probe ToLCNDV-F/ToLCNDV-R/ToLCNDV-P. Samples were analyzed in triplicate. The viral DNA accumulation was estimated by the 2^−ΔΔCt method after qPCR detection of ToLCNDV using the amplification of the COX gene (COX-F/COX-R/COX-P, primer pair and TaqMan probe) as a control to normalize the ToLCNDV cycle threshold (Ct) values. Primers and probes were evaluated in a 10-fold dilution series to establish the standard curve for each, and only reaction efficiencies of ≥93% were considered acceptable, or primer-probe sets were redesigned to achieve an acceptable efficiency. The efficiency of the primer-probe set for quantification of the COX control (Weller et al. 2000) was 94% and the efficiency of the primer-probe to detect ToLCNDV was 93%. Relative quantity (RQ) of viral DNA was calculated by the 2^−ΔΔCt method (Livak and Schmittgen 2001). Fold-change between viral and melon DNA was calculated using the mean Ct of the COX gene as the internal reference for ToLCNDV normalization, and the mean ToLCNDV fold-change of the treatment with the highest viral DNA accumulation as the relative reference for RQ calculations. The qPCR amplification reactions were carried out in a 7300 Fast Real-Time PCR System (Applied Biosystems, Foster City, CA, U.S.A.) under the following conditions: 5 min at 95°C for initial denaturation, followed by 40 cycles of 15 s at 95°C and 30 s at 60°C. The reaction mix (15 µl) consisted of 7.5 µl of 2xSensiFAST PROBE Hi-ROX Mix (Meridian Bioscience Inc.), 10 mM forward and reverse primers, 0.1 mM TaqMan probe, and 65 ng of extracted total DNA as template. Negative controls were obtained from leaves of ToLCNDV-free melon Nst plants. Virus-infected symptomatic leaves of Nst melon plants inoculated with ToLCNDV (the youngest leaf at 15 days postinoculation [dpi]) were used as the positive experimental control for virus detection.

Table 1. Primers used in conventional PCR and primers and TaqMan probes used for quantitative real-time PCR (qPCR) to detect tomato leaf curl New Delhi virus (ToLCNDV), and the melon cytochrome c oxidase (COX) gene in seeds, leaves, and flowers of melon plants

Primers and probesa	Nucleotide sequence (5′ → 3′)	Nucleotide positione	GenBank accession numberf	Reference	Dye
ToLCNDV-A-Fb	AGCACAGCCACGGTGAAGAAC	768–788	KF891468	Ruiz et al. 2015	−
ToLCNDV-A-Rb	TTTCATCCTTCGACAGAGTTC	2,027–2,007	KF891468	Ruiz et al. 2015	−
ToLCNDV-Fb,c	CGTCCTACTATCGCCCTCTATGA	1,917–1,939	U15015	This work	−
ToLCNDV-Rb,c	ATTTCATCCTTCGACAGAGTTCC	2,014–2,036	U15015	This work	−
ToLCNDV-P	ACGACACCCAATCGACGAG	1,988–2,006	U15015	This work	FAM
COX-Fd	CGTCGCATTCCAGATTATCCA	1,488–1,508	X83206	Weller et al. 2000	−
COX-Rd	CAACTACGGATATATAAGAGCCAAAACTG	1,566–1,538	X83206	Weller et al. 2000	−
COX-P	TGCTTACGCTGGATGGAATGCCCT	1,511–1,534	X83206	Weller et al. 2000	YAKYE

^a F = forward, R = reverse, and P = probe.

^b Primers used in conventional PCR.

^c Primers used in qPCR.

^d Primers used in conventional PCR and qPCR to amplify a plant endogenous gene.

^e Nucleotide numbers according to the sequence deposited in GenBank with accession number indicated.

^f GenBank accession number of the sequence used to design the primer.

View as image HTML

ToLCNDV presence in melon flower tissues

Male and hermaphrodite melon flowers at anthesis from ToLCNDV experimentally infected and virus-free melon plants were sampled at 32 dpi and dissected with a scalpel in petals + sepals, stamen (anthers + filament), and pistil (ovary + style + stigma). Total DNA extraction was performed on each of the tissues and the ToLCNDV presence was analyzed by conventional PCR amplification. The detection of endogenous COX gene was used in each sample as an amplification control. Total DNA from the youngest leaf of ToLCNDV-infected Nst melon plants at 15 dpi was used as a positive control for viral detection.

ToLCNDV presence in melon seeds

Mature seeds collected from the fruit of melon plants experimentally infected with ToLCNDV were analyzed by conventional PCR or qPCR using total DNA extracted from seed samples, as described above. Also, seeds were carefully dissected with a sharp scalpel to separate the fractions corresponding to the seed coat and the seed internal tissues (cotyledons and embryo) to analyze the inner or outer localization of ToLCNDV. Seeds from ToLCNDV-free Nst melon plants were used as a negative control. As positive controls, total DNA extracted from the youngest leaf of ToLCNDV-infected Nst melon plants at 15 dpi diluted (1:10 dilution) in total DNA extract from seeds of ToLCNDV-free Nst melon plants was used.

Evaluation of ToLCNDV transmission from melon seed to progeny plant

To investigate the possible transmission of ToLCNDV from the melon seeds to the derived progeny plants, grow-out experiments were conducted. Early growth stage studies were conducted with NT seeds collected from ToLCNDV-infected Nst melon plants that were germinated in transparent plastic boxes (18 by 15 by 7 cm) with filter paper moistened in distilled water. The boxes were placed in germination growth chambers with a photoperiod of 16 and 8 h at 27 and 22°C (day and night, respectively). Plantlets were grown until the first-leaf stage (approximately 15 days); then, cotyledons, first true leaf, and seed coats were individually collected for total DNA extraction and detection of ToLCNDV by qPCR. Equivalent DNA extracts were obtained from plantlets grown in the same conditions from seeds collected from ToLCNDV-free Nst melon plants as control.

Grow-out studies also were conducted with plants at late growth stages. These studies were conducted in two geographically distant locations: Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria (INIA; Madrid), and Instituto de Hortofruticultura Subtropical y Mediterránea “La Mayora” (IHSM; Málaga). In both locations, the assays were performed in insect-proof growth chambers (see above) using melon plants at the 20-leaf stage. In each assay, 5 to 10 melon plants (BR, Nst, and MY melon cultivars) were agroinoculated with ToLCNDV or mock-inoculated (Nst melon cultivar) as positive and negative controls, respectively. Progeny plants were observed periodically (10- and 20-leaf stage, about 40 and 60 days postplanting, respectively) for the development of characteristic ToLCNDV symptoms. The youngest leaf from each plant was collected at 60 days postplanting to detect ToLCNDV presence by tissue-blot hybridization or by PCR or qPCR amplification. For bulk samples, total DNA was isolated from bulks of one leaf per plant. The virus-infected positive control consisted of a bulk of the youngest leaf collected from each of four mock-inoculated plants and from one of the ToLCNDV-symptomatic melon plants to dilute the excess of template and avoid PCR saturation. The negative control consisted of a bulk of the youngest leaf from each of five mock-inoculated plants.

Statistical analysis

The RQs of ToLCNDV DNA detected by qPCR in treated (WT or WD) and NT seeds of the Nst melon type were estimated and compared statistically. Similarly, the RQs of ToLCNDV DNA in the melon seed fractions (seed coat and seed internal tissues) were also estimated and compared statistically. The analysis was performed on biological samples of 10 seeds or 10 dissected fractions of seeds. Because the RQ values did not exhibit a normal distribution and homogeneous variance, the nonparametric Jonckheere-Terpstra’s Median-test for K independent samples was used to statistically compare the virus RQs of the treatments.

Results

ToLCNDV presence in melon floral tissues

In total, four and six melon male and hermaphrodite flowers, respectively, were analyzed at 32 dpi from Nst melon plants experimentally infected with ToLCNDV and from ToLCNDV-free Nst melon plants. Similar results were obtained in all cases for dissected flower tissues (Fig. 1). ToLCNDV presence in floral tissues, either in petals and sepals, stamen, or pistils, was clearly detected in male and hermaphrodite flowers. In contrast, no PCR product was amplified in equivalent floral tissues of virus-free plants. Therefore, ToLCNDV detection in the flower reproductive tissues suggested that the virus might have access to the melon seeds during maturation.

Download as PowerPoint

Detection of ToLCNDV in melon seeds from infected mother plants

Melon plants of cultivars BR, Nst, and MY experimentally infected with ToLCNDV developed characteristic infection symptoms in the youngest noninoculated leaf within 10 dpi. The virus was detected in the youngest leaf by tissue-blot hybridization and PCR amplification in all of the mother plants that served as the seed source for this study. Mature melon seeds were collected from ripe fruit of each cultivar and processed as described above. The results of the PCR and qPCR amplification assays indicated that the virus was detectable in total DNA isolated from the two bulked seed samples (10 NT seeds per sample) of each melon variety, indicating ToLCNDV presence.

Individual seeds collected from plants of the Nst melon variety infected with ToLCNDV were analyzed by qPCR to better understand the localization of the virus. A reduction in ToLCNDV DNA RQ was observed in WT and WD seeds (statistically significant in the latter case) compared with NT seeds (Fig. 2). Based on the results of qPCR amplification, ToLCNDV was not detected (undetermined Ct value obtained) in the seeds produced by the virus-free NT melon plants included as the negative control. The results obtained suggest a major localization of ToLCNDV externally as contaminant of the seed coats. In seeds disinfected with 1.5% sodium hypochlorite, however, ToLCNDV DNA was still detectable by qPCR (Fig. 2, WD treatment).

Download as PowerPoint

To better understand the internal or external localization of ToLCNDV, WD and NT seeds from ToLCNDV-infected Nst melon plants were dissected. The seed coat and the seed internal (cotyledons plus embryo) fractions were analyzed individually for each dissected seed to detect virus presence in each fraction. Equivalent fractions from seeds of ToLCNDV-free melon plants were included as negative control. Based on the results of the qPCR amplification assay summarized in Figure 3, significantly lower ToLCNDV DNA RQ value was observed in the internal seed fraction (seed cotyledons + embryo) of WD seeds compared with the seed coat fraction. In contrast, no significant differences were observed in the ToLCNDV DNA RQ values detected in the seed coat and the seed internal fractions of NT seeds. The DNA extracts of internal or external fractions dissected from seeds collected from virus-free plants resulted in negative ToLCNDV detection according to the undetermined Ct values obtained in qPCR analysis for these samples.

Download as PowerPoint

Transmission of ToLCNDV from seed to progeny melon plant

“Seedborne” means that the virus is present in the seed but not necessarily that plants derived from such seed will be infected. Thus, this latter aspect was further investigated and seed-to-progeny transmission of ToLCNDV was studied at early and late growth stages of melon plants.

NT seeds from ToLCNDV-infected melon plants, which exhibited the highest ToLCNDV DNA RQ (Figs. 2 and 3), were used in grow-out experiments to maximize opportunities for seed transmission of the virus. Melon seeds from a ToLCNDV-free melon plant of the same cultivar were used as negative control (mock control). Early grow-out experiments were conducted with 20 seeds per condition (from ToLCNDV-infected or ToLCNDV-free Nst melon plants) that were individually germinated in a growth chamber and grown until the first-leaf growth stage (approximately 15 days postsowing [dps]), avoiding contact between individual plants. The results of qPCR amplification assays indicated that ToLCNDV was detectable in total DNA extracts of the seed coats collected from the progeny seedlings of seeds removed from the fruit of ToLCNDV-infected mother plants. However, in these early grow-out studies, negative ToLCNDV DNA detection was observed in any of the DNA extracts from cotyledons or the first true leaf of the progeny seedlings. Negative ToLCNDV DNA detection was also observed in DNA extracts from samples (seed coat, cotyledon, and first leaf) from the progeny seedlings of seeds collected from ToLCNDV-free melon plants.

For late grow-out experiments, ToLCNDV infection was analyzed in progeny plants at 20 and 60 dps. NT seeds from ToLCNDV-infected plants of the three melon cultivars (BR, Nst, MY) were used. Three independent assays were conducted, with a total of 775 plants tested for the three melon cultivars: one at location 1 (INIA) and two at location 2 (IHSM). Visual examination at 20 and 60 dps revealed that progeny plants of seeds collected from ToLNDV-infected plants did not develop symptoms of infection. In contrast, characteristic ToLCNDV infection symptoms were observed in all ToLCNDV-experimentally infected control plants. Results obtained for viral detection in the youngest leaf of progeny plants at 60 dps are summarized in Table 2. Analysis of the progeny plants by hybridization and conventional PCR or qPCR did not show evidence of ToLCNDV infection in any of the 775 plants evaluated. Positive control plants resulted in ToLCNDV detection whereas negative ToLCNDV detection was observed in mock-inoculated control plants.

Table 2. Results of hybridization, PCR, or quantitative real-time PCR (qPCR) analysis to evaluate seed-to-progeny plant transmission of tomato leaf curl New Delhi virus (ToLCNDV) in melon^a

	Assay 1, location 1			Assay 2, location 2				Assay 3, location 2
		qPCR			Hybridization	PCR			Hybridization	PCR
Cultivar	Plantsb	Bulksc	Pos/bulkd	Plantsb	Pos/plantse	Bulksc	Pos/bulkd	Plantsb	Pos/plantse	Bulksc	Pos/bulkd
BR	100	20	0/20	96	0/96	10	0/10	70	0/70	7	0/7
Nst	100	20	0/20	83	0/83	9	0/9	67	0/67	7	0/7
MY	100	20	0/20	94	0/94	10	0/10	65	0/65	7	0/7
BR+f	5	1	1/1	10	9/10	1	1/1	5	4/5	1	1/1
Nst+f	5	1	1/1	10	10/10	1	1/1	5	5/5	1	1/1
MY+f	5	1	1/1	10	10/10	1	1/1	5	5/5	1	1/1
Nst−f	5	1	0/1	10	0/10	1	0/1	10	0/10	1	0/1

^a Progeny plants derived from nontreated seeds collected from fruit of ToLCNDV-infected melon cultivars Brimos (BR), Mayor (MY), or Nesta (Nst) plants were individually analyzed at 60 days postplanting for ToLCNDV presence in the youngest, newly emerged leaf. The analysis was conducted by tissue-blot hybridization of squash blots of leaf petioles with a probe specific to ToLCNDV or by PCR (primer pair ToLCNDV-A-F/ToLCNDV-A-R) or qPCR (primer pair ToLCNDV-F/ToLCNDV-R; probe ToLCNDV-P) analysis of total DNA extracts from bulked leaf samples (5 to 10 leaves per bulk). Assays were conducted in location 1 (Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria, Madrid, Spain) and location 2 (Instituto de Hortofruticultura Subtropical y Mediterránea “La Mayora”, Málaga, Spain).

^b Total number of analyzed plants.

^c Total number of analyzed bulks, from 5 to 10 young leaves per bulk.

^d Number of positive bulks/total number of analyzed bulks.

^e Number of positive plants/total number of analyzed plants.

^f Control plants (BR+, Nst+, and MY+) were experimentally inoculated by Agrobacterium tumefaciens–mediated inoculation using an infectious clone of the isolate ES-Alm-661-Sq-13 of the “Spain” (ES) strain of ToLCNDV (DNA-A and DNA-B, GenBank accession numbers KF749223 and KF749226, respectively) (BR+, Nst+, and MY+), or were mock inoculated (Nst−) (two- to three-leaf stage) and analyzed at equivalent time of plants derived from seeds.

View as image HTML

Discussion

Although seed transmission of plant viruses is not the norm, it can have epidemiological relevance for the dissemination of a virus in nature and can be a key component for its prevalence and perpetuation (Johansen et al. 1994; Lipsitch et al. 1995). That virus is seedborne will not necessarily result in transmission to progeny plants derived from such seeds (Bhat and Rao 2020). No epidemiological implications exist for viruses that are harbored by seed but are not seed transmitted and do not infect progeny plants. However, if transmission through seed occurs, it has epidemiological consequences: first, providing a means for long-distance viral dissemination via commercial seed trade, and, second, resulting in primary infection foci for secondary spread in viral disease outbreaks (Campbell and Madden 1990; Maule and Wang 1996). For seedborne viruses, therefore, the passage of the virus from the seeds to the progeny plants can have significant importance if such plants are of epidemiological significance. This could be particularly relevant for begomoviruses in which effective secondary spread by the insect vector B. tabaci occurs from infected plants (Gilbertson et al. 2015). In this work, we studied the possible transmission through melon seeds to offspring of an isolate of the ES strain of ToLCNDV.

The ability of a virus to move within the host and to accumulate in the reproductive tissues are major determinants of virus seed transmission (Cobos et al. 2019). Here, the ability of ToLCNDV to associate with flower tissues in melon plants was shown. The virus was detected in petal and sepal, stamen, and pistil tissues, supporting its possible association with seeds. In fact, the virus was readily detected in mature seeds collected from infected plants of three melon cultivars, demonstrating that ToLCNDV is seedborne but not seed transmissible (see below). The significant decrease in ToLCNDV DNA relative amount detected in surface-disinfected seeds is compatible with a major association of the virus as an external contaminant of the melon seed coats (Córdoba-Sellés et al. 2007; Davino et al. 2020). Begomoviral detection in tomato seeds has similarly been reported for TYLCV, and surface disinfection was reported to decrease virus DNA accumulation (Pérez-Padilla et al. 2020). Nevertheless, small ToLCNDV DNA RQ was still detected in surface-disinfected melon seeds, which could indicate that some virus was still present, associated externally with disinfected seeds, or that the virus could also be located in internal seed tissues. The analyses of dissected melon seeds conducted here provided additional evidence supporting major localization of ToLCNDV externally in melon seed coats. However, the presence of the virus infecting seed cotyledons or embryo cannot be ruled out, because a low level of ToLCNDV DNA was detectable in the internal fraction of the disinfected melon seeds. The occurrence of cross-contaminations during seed dissection with viral DNA present in seed coats cannot be ruled out. In fact, cross-contamination of dissected seed internal fractions with ToLCNDV DNA located externally in the seed coat could explain why no significant difference was observed for NT seeds in the relative virus DNA quantity detected in seed coat versus seed internal fractions.

Recently, begomovirus seed transmission has been reported for several plant species (Fadhila et al. 2020; Kil et al. 2016, 2018; Suruthi et al. 2018) comprising species belonging to the Cucurbitaceae family such as for isolates of ToLCNDV in chayote (Sechium edule (Jacq.) Sw.) (Sangeetha et al. 2018) or for bitter gourd yellow mosaic virus in bitter gourd (Momordica charantia L.).(Manivannan et al. 2019). In chayote, seed material planted are fruit with in situ germinating seeds, thus being a special case of seed transmission. Seed transmission of an isolate of the ES strain of ToLCNDV has been reported in zucchini squash at a rate of 61.36% (Kil et al. 2020). However, the results of the studies of Sangeetha et al. (2018), Manivannan et al. (2019), and Kil et al. (2020) did not demonstrate whether plants resulting from seed shown to harbor virus in laboratory assays did serve as a source of virus inoculum for subsequent transmission by the whitefly vector.

The ToLCNDV detection in melon seeds observed here might suggest the possible vertical transmission of the virus to the offspring, as observed for other plant viruses (Cobos et al. 2019). However, the results from the specific grow-out experiments conducted in the current work demonstrated that, in more than 800 plants tested from seeds collected from ToLCNDV-infected plants of the three melon cultivars, ToLCNDV was never detected in progeny melon plants at either early or late growth stages. Therefore, although ToLCNDV was demonstrated here to be seedborne in melon, the results did not support seed transmission to emerging seedlings. Equivalent results were also obtained recently for isolates of Sweet potato leaf curl virus, another begomovirus species, which was found to be seedborne in sweet potato but not seed transmitted to emerging seedlings (Andreason et al. 2021). Similarly, seed transmission was discarded recently in tomato plants for other begomoviruses such as isolates of the species Tomato yellow leaf curl virus and Tomato yellow leaf curl Sardinia virus (Pérez-Padilla et al. 2020; Tabein et al. 2021).

In general, seed transmission of viruses is unusual and only stable viruses such as tobamoviruses remain biologically active for transmission when carried as contaminants in seeds (Johansen et al. 1994; Jones 2021; Salem et al. 2022). For most seedborne viruses, the virus can be present in immature seeds but is inactivated in both seed coat and embryo during maturation (Bhat and Rao 2020; Bowers and Goodman 1979). In fact, as indicated by Maule and Wang (1996), only a few viruses are sufficiently stable to survive during seed formation and to withstand exposure to seed dehydrations, harvest, and storage; and, if present, viable viruses rarely contact and infect the seedling. Invasion of seeds, even of seed embryos, has been demonstrated for several begomoviruses, leading to putative seed transmission (Kim et al. 2015; Kothandaraman et al. 2016; Manivannan et al. 2019). However, even when detected in seeds, absence of seed transmission was also observed for other begomovirus–host combinations (Andreason et al. 2021; Pérez-Padilla et al. 2020; Rosas-Díaz et al. 2017; Sisodia and Mahatma 2020). Thus, the fact that some begomoviruses are seedborne is without question and has been demonstrated in various scientific papers. Nevertheless, seed transmission, understood as transmission to progeny plants resulting in virus sources for the whitefly vector for subsequent spread at epidemiological significance, generally remains unresolved. The results of the current work and equivalent studies (Andreason et al. 2021; Pérez-Padilla et al. 2020; Rosas-Díaz et al. 2017; Tabein et al. 2021) differed from the high seed transmission rates reported for begomoviruses by Kil et al. (2016, 2020). However, experimental contamination cannot be excluded in the latter studies because plants from seeds collected from healthy plants were not included as control.

The epidemiological importance of begomoviruses that, at times and under certain circumstances, are detectable in mature seed of certain vegetable species and cultivars and the progeny plants has not been fully assessed. Until now, whitefly vector transmission has been the only epidemiologically significant mode of begomovirus transmission. If seed transmission was a prominent mode of transmission among begomoviruses, previous field or greenhouse observations would have more than likely pointed to the potential for alternative modes of transmission. Exceptions might be found among certain cultivars planted for seed production in particular environments over the others, making this a reasonable and logical point to address in future research to define the conditions that may be conducive to viruses being harbored on or in the seed and eliminate those conditions. Epidemiological information available until now supports the paradigm that transmission of begomoviruses by seed is the exception, not the norm. In addition, highly sensitive molecular assays may also lead to erroneous conclusions if experimental designs are flawed, or aerosol contamination of laboratory equipment (including pipettes and tubes, among others) cause distorted results, which can culminate in misleading inference.

In conclusion, no evidence was found for ToLCNDV seed transmission in melon for the three melon varieties tested. However, viral DNA could be detected in mature seeds collected from ToLCNDV-infected melon plants, suggesting that the virus is seedborne but effective transmission to progeny seedlings does not occur. Nevertheless, the possible existence of specific interactions between ToLCNDV isolates and host plants that could favor seed transmissibility cannot be ruled out. Because differences in the ability of virus isolates to invade plant tissues have been reported (Johansen et al. 1994; Maule and Wang 1996; Morra and Petty 2000), it is not unexpected that such differences could result in cultivar-to-cultivar variation that could influence invasion of the seed and seed parts by viruses. However, studies providing definitive evidence and outcomes are far from comprehensive across plant virus genera and families, making epidemiological observations and family-wide patterns of transmission modes relevant evidence for construing modes of virus spread in crops to guide effective management strategies.

Acknowledgments

We thank S. Fitó for kindly providing melon seeds.