Introduction

Diseases produced by viruses cause economic losses in commercial cucurbit production around the world (Lovisolo 1980). Among these viruses, Cucumber green mottle mosaic virus (CGMMV) represents a major risk in the production of melon, watermelon and cucumber. CGMMV belongs to the genus Tobamovirus, family Virgaviridae (Adams et al. 2009) and causes systemic mottle and mosaic symptoms on cucurbitaceous plant leaves (Kim et al. 2003). In watermelon, the virus causes a pulp deterioration called blood flesh disease (Mandal et al. 2008) and the fruit loses its marketable value. In cucumber, the virus causes deformation and mosaic symptoms (van Koot and van Dorst 1959). It is mechanically, pollen- and seed-transmitted (Liu et al. 2014).

CGMMV was first described infecting cucumber in England by Ainsworth (1935). Its incidence in other countries of the world has increased rapidly during the last decade, possibly through the international seed trade following cucurbit seed crop production in tropical or subtropical countries (Dombrovsky et al. 2017). Also, seed-testing routines for CGMMV may be inadequate, which allows for the rapid and worldwide spread of the virus. Contaminated seeds provide a route for the movement of the virus between countries and its introduction into new areas, and several seed treatments currently used, were found insufficient to eliminate the virus from contaminated seed lots. In addition, like other tobamoviruses, CGMMV can survive for a long time on plant debris from infected crops (Reingold et al. 2015). Therefore, control depends on early monitoring, awareness of the farmers, and an appropriate crop management, but even implementing these measures, success is not guaranteed. (Reingold et al. 2016).

Worldwide CGMMV isolates are grouped in two major clusters based on biological differences and genome sequences: A first cluster (I) constitutes the European strain and includes most isolates from France, the Netherlands, and Uzbekistan. A second cluster (II) is formed by isolates from Asian countries such as Japan and South Korea (Crespo et al. 2017). Spain is currently the first country where both these strains have been described co-infecting the same crops and in the same region (Crespo et al. 2017).

Commercial tomato and pepper hybrids that carry virus-resistance genes successfully control tobamoviruses in solanaceous crops. But in the case of CGMMV, there is an urgent need for resistant cultivars in cucumber, with restricted virus movement, replication and symptom development. However, currently cucumber varieties resistant to CGMMV are hardly available. Efforts have been made to produce transgenic resistance in the experimental plant Nicotiana benthamiana (Kamachi et al. 2007), in watermelon (Park et al. 2005; Lin et al. 2012) and in melon (Ali et al. 2012). Alternatively, recent reports showed that Stenotrophomonas maltophilia HWS exhibits a good biocontrol against the virus by increasing the expression of defense response genes (Li et al. 2016) and an attenuated strain used as a biocontrol agent reduced yield losses in cucumber (Slavokhotova, et al. 2016).

The development of resistant varieties through conventional breeding could offer a good solution to this disease which continues to escalate. Resistance to CGMMV has been reported only in melon and in wild Cucumis species (Rajamony et al. 1987, 1990a, b). In Cucumis, several wild species of African origin are classified as resistant. One of them is C. anguria L., which has one dominant gene that confers resistance to CGMMV (den Nijs 1982). C. figarei has been described as immune, while other resistant types like snapmelon “Kachri” and other Cucumis sp. are symptomless carriers (Pan and More 1996). In melon, a new source of resistance to CGMMV-SH has been identified in accession “Chang bougi” (Sugiyama et al. 2006). In cucumber, C. sativus var. hardwickii (R.) Alef. has been used to introduce resistance to CGMMV (Carnide and Barroso 2006).

In the present paper we used a real-time RT-PCR test to determine the viral loads of the European and the Asian strain of CGMMV. An isolate of each strain was mechanically inoculated in a collection of 56 C. sativus accessions. We report on the evolution of symptom expression for both virus strains in all accessions, and that of the viral loads on a representative number of plants.

Materials and methods

Plant material

Accessions from a C. sativus collection were supplied by the Institute for Conservation and Improvement of Valencian Agrodiversity (COMAV-UPV) representing three variants of the species: 50 accessions of C. sativus L., 5 of C. sativus var. hardwickii (Royle) Gabaev and 1 of C. sativus var. sikkimensis (Hook.f.). Some of the accessions were collected by COMAV and others were obtained from exchanges with other germplasm banks (USDA-NPGS and CATIE). The selected accessions were meant to represent the variability of the full collection. In addition, we tested C. anguria L. and C. metuliferus E. Mey as African wild species and which were supplied by Seednative (La Iruela, Spain).

Virus sources for mechanical inoculation

In order to investigate differences in resistance, we mechanically inoculated plants with one of two isolates that represent the two different strains of CGMMV: the strain of European origin (CGMMV-SP) (GeneBank GQ411361) and that of Asian origin (CG-SPCu16), the latter as determined by Crespo et al. (2017). Although both strains differ in their genome sequences, they do not show any biological difference in terms of systemic symptoms expressed on leaves of infected cucumber. Both strains showed severe mosaic symptoms and blisters. However, they behaved differently in Chenopodium amaranticolor in which local lesions appeared when Asian-like strains were inoculated (Crespo et al. 2017). Both strains were isolated at IFAPA from cucumber crops in Almeria (Spain) in 2009 and 2015, respectively. They had been isolated after three passages into Gomphrena globosa L., then propagated in cucumber (cv. Cumlaude), and stored at − 80 °C until used for the mechanical inoculation. Before use in screening the cucumber accessions, the isolates were propagated after mechanical inoculation in cucumber (cv. Cumlaude). Infected plants were grown in an insect-proof greenhouse where temperature was partially controlled (25–30 °C). Approximately 3 weeks after inoculation, plants that showed typical symptoms of virus infection were used as the virus source.

Mechanical inoculation

For each virus isolate, 0.5 g of tissue was taken, 5 weeks after sowing and from the second leaf down from the plant apex which displayed CGMMV symptoms. The tissue was homogenized in 1.5 ml of 50 mM sodium phosphate buffer (pH 7.0) and inoculated mechanically by rubbing 150 µl of the extract onto leaves dusted with carborundum powder. At least ten plants of each species were inoculated with each virus isolate and the experiment was repeated twice, each time with two repetitions, during the winter (January–March) and in spring–summer period (May–July) of 2017. All inoculated plants were maintained in an insect-proof greenhouse under controlled conditions.

Evaluation of symptoms and detection of the virus

Inoculated plants were evaluated for the expression of CGMMV symptoms at 23, 33 and 47 days post inoculation (dpi), using the following scale: 0 (symptomless), 1 (mild symptoms, as initial mottle mosaic on leaves), 2 (intermediate symptoms, as evident leaf mottle mosaic on leaves) and 3 (severe symptoms as mottle mosaic, interveinal chlorosis, and blistering in leaves, and distortion and mosaic in fruits). The presence of the virus was analyzed at 15 dpi using a conventional RT-PCR reaction using the primers that amplify part of the coat protein-coding region as described in Crespo et al. (2017).

From plants of selected accessions, 0.2 g tissue was removed from the second leaf (not inoculated, and possibly representing systemic infection) from the apex at 23 (plants with 3 new leaves formed), 33 (plants with 5 new leaves) and 47 (plants with 8 new leaves) dpi for analysis by real-time RT-PCR which was developed for the detection and quantification of CGMMV. The infected leaf tissue of each plant was ground to a fine powder in liquid nitrogen in a pestle and mortar and placed in a sterile microcentrifugue tube. Total RNA was extracted with Trizol reagent (Invitrogen). The resulting pellet was resuspended in 50 µl DEPC-treated water and stored at − 80 °C. RNA was quantified with a ND-2000c Spectrophotometer (NanoDrop Technologies) and diluted to a final concentration of 50 ng/µl. Real time RT-PCR reactions were set up in 96-well reaction plates using TaqMan One Step RT-PCR Master Mix (Applied Biosystems). One microlitre aliquots were used as templates in the RT-PCR reactions of 20 µl, containing 10 µl Master Mix, 1 µl forward primer, 1 µl reverse primer, 0.5 µl probe, 1 µl total RNA and 6.5 µl DEPC-water. Following Chen et al. (2008), the primers and probe for the TaqMan assay were 5′-GCATAGTGCTTTCCCGTTCAC-3′ (sense) at positions 6285–6305nt and TGCAGAATTACTGCCCATAGAAAC-3′ (antisense) at positions 6362–6385nt. The probe was 5′-CGGTTTGCTCATTGGTTTGCGGA-3′ at positions 6316–6338nt, labeled with 6-carboxyfluorescein (FAM) and the 3′end was labeled with N,N,N′,N′-tetramethyl-6-carboxyrhodamine (TAMRA). The primers and probe for amplification of the internal controls were: primers CUC18S-For (5′-GGCGGATGTTGCTTTAAGGA-3′) and CUC18S-Rev (5′-GTGGTGCCCTTCCGTCAAT-3′; probe CUC18S-Ana (5′-TCCGCCAGCACCTTATGAGAAATCAAAGTC-3′) labeled with JOE (2,7-dimethoxy-4,5-dichloro-6-carboxyfluorescein) as the 5′terminal reporter dye and with BHQ1 (Black Hole Quencher One) as the 3′ quencher dye (Gil-Salas et al. 2009). Three technical replications were performed per sample and the tests were run on ABI Prism 7000 DNA sequence detection system (Applied Biosystems) as follow: 10 min at 50 °C, 1 min at 95 °C, 40 cycles of 15 s at 95 °C and 1 min at 60 °C.

Relative accumulation of CGMMV in the plants was calculated by the comparative Ct (cycle threshold) method, using as an endogenous reference gene a fragment within the 18S ribosomal region of the cucurbit mitochondrial genome (Gil-Salas et al. 2009), and as a calibrator, untreated control (uninfected cucumber). In this way, with RT-PCR we obtained amplified fragments of the expected size using real-time PCR primers (virus and internal control). The expected PCR products were cloned and sequenced. The sequence identity of the internal control and the virus was confirmed after comparison with GQ411361 and AF206894 (C. sativus 18S ribosomal RNA gene) GenBank sequences. The relative accumulation of the virus from collected samples was calculated using the 2−ΔΔCt formula. Here, ΔCt is the difference between the Ct of each sample (mean of the three technical repetitions) and the Ct of the internal control (mean of the three technical repetitions) (ΔCt = Ct sample − Ct internal control), and ΔΔCt is the difference between ΔCt and ΔCt calibrator (ΔΔCt = ΔCt − ΔCt calibrator). Finally, 2−ΔΔCt represents the amount of target, normalized to the endogenous reference and relative to the calibrator (Livak and Schmittgen 2001; Sáez et al. 2016).

Data analysis

Resistance was evaluated as the response of the host plant to virus infection estimated from symptom severity in all inoculated plants, and from the viral titer in a selected number of genotypes (12 accessions representing C. sativus; C. anguria and C. metuliferus). The interaction effects from inoculated strain of virus, dpi and symptoms were investigated using general linear model statistics. All effects were tested at the 5% significance level. Analyses were performed using Statistics 9.1 statistical software.

Results

Symptom expression

Single inoculations in cucumber with CGMMV isolates CGMMV-SP and CG-SPCu16 that represent the European and the Asian strain, respectively, produced three clearly differentiated levels of symptoms: mild symptoms (soft mottling), intermediate symptoms (limited to evident mottle-mosaic on leaves), or severe symptoms, that showed as leaf mottle mosaic, interveinal chlorosis, and blistering in leaves, with distortion and mosaic in fruits (Fig. 1). The time of appearance of these symptoms in individual plants was variable, but all of the replicate plants of each accesion showed similar symptoms at 23 dpi. We observed no change in the symptoms at 33 and 47 dpi. Most of the 56 tested C. sativus accessions were highly sensitive to the mechanical transmission of CGMMV, 44 of which developing severe symptoms including C. sativus var. hardwickii. Only 2 accessions showed mild symptoms (BGV001358, from Calcuta, and CGN19818 from Dzampur, India) and 10 accessions from India and Spain showed intermediate symptoms up to the end of the test period, including C. sativus var. sikkimensis (Table 1). Finally, regarding to the African wild species, C. anguria was symptomless and tested negative in conventional RT-PCR, whereas C. metuliferus showed severe symptoms and tested positive (Fig. 2).

Fig. 1
figure 1

Leaf symptoms on cucumber at 23 days post-inoculation with CGMMV-SP: (a) mild symptoms, (b) intermediate symptoms, (c) severe symptoms

Full size image
Table 1 Response of Cucumis sativus accessions, C. anguria and C. metuliferus, to mechanical inoculation with CGMMV
Full size table
Fig. 2
figure 2

C. anguria (a), without visible symptoms, and C. metuliferus (b), showing severe symptoms at 23 days post-inoculation with CGMMV-SP. Identical results were obtained following inoculation with CG-SpCu16

Full size image

Identical results were obtained in two repeat experiments under the same conditions and following the methodology described above, one during the winter period (January–March 2016–2017) and the second one in the spring–summer period (May–July 2017).

CGMMV viral loads

All inoculated plants, except for C. anguria, tested positive by the conventional RT-PCR test. Plants from a selection of accessions (marked in Table 1) that covered the three types of symptoms when inoculated with the CGMMV strains, were analyzed with real-time RT-PCR in samples collected at 23, 33 and 47 dpi (see Table as Supplementary Material). The accessions selected to represent the range of symptoms, consisted of 5 genotypes that showed severe, 5 that showed intermediate, and 2 that showed mild symptoms. The results from the accumulated analysis of variance on the real-time RT-PCR data showed a significant effect with respect to the level of expressed symptoms (P < 0.001). This suggested that the viral titer was positively correlated with the degree of symptoms expressed in the inoculated plants. Moreover, the difference in the normalized amount of CGMMV (log 2−ΔΔCt) when the cucumbers had been inoculated with European or Asian genotypes was also statistically very significant (P < 0.001; Table 2). In addition, the amounts of virus of CG-SPCu16 (Asian strain) were higher in accessions with intermediate symptoms at 23 and at 33 dpi (P < 0.001 and P < 0.05, respectively), when compared with those infected with CGMMV-SP (Fig. 3). On the other hand, the amounts of virus detected over time (dpi) regarding to the symptom observed, did not change significantly (Table 2).

Table 2 Accumulated analysis of variance results of the normalised amount (log 2−ΔΔCT) of CGMMV
Full size table
Fig. 3
figure 3

Error-Bar plots of normalized amounts of CGMMV (log 2 − ΔΔCT) for the real-time RT-PCR analysis in selected cucumber accessions (marked in Table 2) for days post inoculation (dpi) (a: at 23 dpi, b: at 33 dpi or c: at 47 dpi), symptom (1: mild, 2: intermediate or 3: severe), and virus isolate (SP: from European origin, CGMMV-SP, or 16: from Asian origin, CG-SPCu16); *P < 0.01

Full size image

Discussion

We tested a collection of cucumber germplasm accessions against CGGMV, and most genotypes were very sensitive to the virus after mechanically inoculation. Moreover, the high susceptibility observed along the whole range of diversity of this species confirmed that this virus represents a major threat to cucumber cultivation. Out of 58 accessions evaluated, 10 produced only intermediate level symptoms, suggesting a certain level of tolerance. Interestingly, we also have identified 2 accessions (BGV001358 and CGN19818) that showed only mild symptoms upon infection with CGMMV, and that accumulated a low viral titer. Resistance is an important factor that determines concentrations of virus in several virus-cucurbita host plant pathosystems. This happens in the case of Cucurbit yellow stunting disorder virus and Watermelon mosaic virus in melon (Marco et al. 2003; Díaz-Pendón et al. 2005), and Papaya ringspot virus in squash and watermelon (Pacheco et al. 2003). The plants from the two accessions, both originated from India, produced mild symptoms after inoculation with CGMMV, and may have a higher level of resistance. This resistance could be related with the co-evolution of host and pathogen in this part of the world.

Different cultivars may accumulate different amounts of virus when infected with CGMMV, and those plants that have restricted viral load could be regarded as potential sources of resistance (Cech and Branisovâ 1976). Other reports of CGMMV resistance tests in vegetable accessions have enabled to select only a limited number of interesting cultivars. Of 345 cultivars tested against CGMMV, all produced mosaic symptoms in leaves, except for only one, named “Hanboksamcheok”, which showed mild mosaic symptom in a field test (Ko et al. 2004).

C. metuliferus was found very susceptible to CGMMV (Table 1, Fig. 2). In contrast, the wild species C. anguria could be a potential source of resistance to CGMMV due to the fact that it did not develop any symptoms following inoculation and that it remained negative for the virus when tested with real-time RT-PCR, suggesting that this species is not susceptible to CGMMV. Although attempts to cross C. anguria with either cucumber or melon have failed so far (Deakin et al. 1971; Kho et al. 1980; Kroon et al. 1979), there appears to be a possibility that some kind of hybridization can be achieved between C. sativus and C. anguria (den Nijs 1982).

In many virus-host plant pathosystems, concentrations of virus in infected leaves have been found to rise rapidly and subsequently decline (Matthews 1991). Cucumber plants inoculated with Cucumber vein yellowing virus accumulated concentrations of the virus during 2 months, but some cultivars showed reductions in concentrations sampled at 45 dpi coinciding with the time of fructification, during which fruits exerts a significant sink demand and become highly competitive with vegetative growth (Pharr et al. 1985; Gil-Salas, et al. 2009). As already suggested over 40 years ago, tolerance to CGMMV may be associated with differences in concentration of virus, although there need not be a direct relationship between virus symptoms and the virus concentration. Therefore, the evaluation of resistance against tobamoviruses should include symptomatic and quantitative virus assays (Cech and Branisovâ 1976). In the present study we found a significant relationship between the expressed symptoms and the concentrations of CGMMV-SP or CG-SpCu16. Furthermore, some plants with very mild symptoms contained significant amounts of the virus, which could enhance the possibilities of further spread in crops. In contrast to what is observed with other cucurbit viruses (Gil-Salas et al. 2009), CGMMV virus load did not change significantly over time (Table 2). In this case, we have not found a relationship between fructification or flowering and a reduction in virus titer.

Taqman real-time RT-PCR has been developed for routine detection of CGMMV and is recommended for use in the field of plant quarantine; this technique, as described by Chen et al. (2008), detects low quantities of virus which would be assist in the prevention and control of the disease caused by this virus. In the present paper we supplied the qRT-PCR test with an internal control based on the cucumber 18S house-keeping gene to normalize the quantification of the viral sequence (Gil-Salas et al. 2009), so that this technique can be used in cucumber breeding programs to estimate the amounts of CGMMV in inoculated plants. We found that the CGMMV-specific primers and probe were able to detect isolates from the European as well as the Asian strain. These strains are genetically and biologically different, yet symptomatology in the field can be similar, and both can be present in the same fields, especially after the recent introduction of the Asian-type CGMMV in many countries of the world (Crespo et al. 2017; Dombrovsky et al. 2017). We found that the overall concentrations of both virus strains in cucumber were different (P < 0.001; Table 2). They were similar in plants that were most sensitive and those that showed mild symptoms (P > 0.05 in all cases). However, in plants that showed intermediate symptoms, the viral titers of CG-SpCu16 were significantly higher rather than those of CGMMV-SP at 23 and 33 dpi (Fig. 3), which may be explained by the differences between the genome sequences as well as the different reactions in determined host plants following inoculation from both strains as has been previously described (Crespo et al. 2017). This should be taken into account during plant breeding programs, and the susceptibility and sensitivity of candidate accessions should be checked with both strains which have different geographic origins. High seed transmission rates (over 75%) from CGMMV-infected cucumber plants have been reported, and the seed-borne nature of this virus is considered the main cause of spread to new countries and areas (Liu et al. 2014). However, CGMMV is found both on and within the seed coat, so detection in seed batches does not differentiate CGMMV that is contaminating seeds externally from that which will infect the seedling (Reingold et al. 2015). The differential viral accumulation in sensitive and resistant cucumber accessions and differential accumulation of European and Asian strains could also now be investigated with respect to the presence of CGMMV in seeds and its relevance to the risk of transmission in the field.