Nepovirus nicotianae (Tobacco ringspot virus; TRSV) is assigned to the genus Nepovirus, subgroup A, consisting of two single-stranded RNA segments that both contain a single open reading frame (ORF) encoding for a polyprotein (Fuchs et al., 2022). It has a wide host range, including many ornamental species. TRSV is listed as a quarantine pest in the European Union (EU). To qualify for listing as a quarantine pest, the identity of the pest has to be established, the pest needs to be absent in the regulated region, the pest should be able to become established or spread in the regulated area and have potential economic impact (FAO, 2013). TRSV was regulated because major losses have been reported in blueberry (Vaccinium corymbosum), grapevine (Vitis vinifera) and soybean (Glycine max) productions (EFSA PLH Panel et al., 2019). Although quarantine pests are supposed to be absent in the EU, TRSV has been reported worldwide, also in five EU member states: Czech Republic (Kundu et al., 2016), Italy (Bellardi & Marani, 1985; Sorrentino et al., 2012), Lithuania (Samuitienė & Navalinskienė, 2001), the Netherlands (Asjes, 1974), Poland (Kaminska, 1985), and the former member state United Kingdom (UK) (DEFRA, 2018; Stone, 1980). The virus is naturally spread by five nematode species belonging to the Xiphinema americanum sensu lato (s.l.) group, which occur around the world, but from these vector species only X. rivesi has been reported in a few EU member states (EFSA PLH Panel et al., 2018; EPPO, 2022b). Although transmission by European populations of X. rivesi has been shown under experimental conditions (Sirca et al., 2007), natural spread of TRSV has not been reported in the EU, despite occasional findings of TRSV.

EU member states are required to perform surveys to guarantee the absence of quarantine pathogens in their territories. The previously mentioned reports of TRSV in ornamentals, together with findings of TRSV in 1997 and 2000 in ornamentals in the Netherlands, showed that TRSV could remain symptomless in various ornamental species. As a consequence TRSV might have gone unnoticed in various hosts in the EU including the Netherlands. This prompted the Netherlands Institute for Vectors, Invasive plants and Plant health (NIVIP) and the Netherlands Inspection Service for Horticulture (Naktuinbouw) to initiate nationwide survey programs in vegetatively propagated ornamentals for TRSV and in soil for its vector species. Here we report the results of these surveys and other incidental TRSV findings since 1997. Additionally, the results of a phylogenetic analysis of (near) complete genome sequences of available TRSV isolates from these findings and from the NCBI GenBank are discussed in a phytosanitary context.

In 2005, a TRSV survey program was initiated, focusing on ornamental species (Table 1). A species selection was made based on previous findings, reports in literature and notifications from other countries. Samples were taken regardless of symptoms and consisted of leaf material from five plants per lot, i.e. a trade consignment consisting of one cultivar. For Ajuga reptans, Hemerocallis, Iris germanica, and Phlox subulata (2017–2020), samples consisted of leaf material from 200 plants per lot, tested in subsamples from 5 plants. For the bulbous crops Lillium and Tulipa, root material was sampled and tested because leaf material wasn’t always available at the time of sampling. Samples for both surveys and other findings consisted of about 1 g leaf or root material. Additionally, in 2013, a survey program for X. americanum s.l. and other Xiphinema species was initiated by collecting soil samples from randomly selected orchards, fruit tree nurseries and flower-bulb fields. Furthermore, soil samples were taken from fields of growers where TRSV was identified in Hemerocallis, Iris ensata, and I. germanica in 2006, 2007 and 2018 (Table 2).

Table 1 List of ornamental plants screened for presence of TRSV during official surveys in the Netherlands from 2005–2020, including number of samples per year
Full size table
Table 2 List of site types screened for presence of Xiphinema species in soil during official surveys in the Netherlands from 2006–2021, including number of samples per site type, and what Xiphinema species was identified
Full size table

Since the reported findings span 20 years, it is important to note that the number of available detection and identification methods have increase over the years, as has been described by (Fox & Mumford, 2017). This means that different techniques were used in different time periods when testing for TRSV in The Netherlands. Before molecular methods were available, samples were tested using bioassay and DAS-ELISA. Since 2017, Illumina sequencing was used to determine full genome sequences after TRSV detection by bioassay or DAS-ELISA. In 2019 the use of real-time RT-PCR got a prominent role in detection, while bioassay and DAS-ELISA were used to support both molecular methods. From seven of the ten TRSV findings before 2017, an isolate was maintained in frozen plant material at NIVIP. These isolates were analyzed with Illumina sequencing in 2022 and were included in the validation of the real-time RT-PCR.

Bioassays were performed by inoculating plant sap from leaf or root tissue to a range of test plants routinely used as indicators in virus diagnosis: Chenopodium quinoa, Nicotiana benthamiana and Nicotiana occidentalis P1 (EPPO, 2022a; Verhoeven & Roenhorst, 2000). DAS-ELISA was either performed on leaf or root material of the original host or on symptomatic leaf material from inoculated plants using the TRSV antiserum from Prime Diagnostics (Wageningen, The Netherlands) (EPPO, 2015).

A real-time RT-PCR specific for detection of TRSV was designed in 2019, based on in silico analyses of GenBank accessions (see Supplementary material): primers TRSV-F1 (TGT TAA TGT CCA TTT TTG GTA GAA ART G), TRSV-F2 (TGT TGA TGT CTA TTT TTG GTA GAA AGT G), TRSV-R (TAA GTG CAC ACC CCG ACG G), and TRSV-P (FAM-TCC CAA GTT TTT GAA GTT CGG GGT GG-BHQ1). After grinding leaf material in GH + buffer (Botermans et al., 2013), RNA was isolated, using the Sbeadex® maxi plant kit (LGC genomics) on a KingFisher KF96 system following the manufacturer’s instructions. Reaction mixtures consisted of 6.25 µl UltraPlex™ 1-Step ToughMix® (4x) (Quanta Biosciences, Plain City, Ohio, USA), 0.15 µM of each forward primer, 0.3 µM of the reverse primer, 0.1 µM of the probe, 2 µl RNA extract, an internal RNA isolation control (nad5) (Botermans et al., 2013; Menzel et al., 2002), i.e., 0.02 µM of each primer and 0.1 µM of the probe, and RNAse-free water added up to a final volume of 25 µl. Real-time RT-PCR was performed in 96-well plates on a Bio-Rad CFX96™ real-time PCR Detection System (Bio-Rad Laboratories Hercules, California, USA): 10 min 50 °C, 3 min 95 °C, 40 cycles 10 s 95 °C and 1 min 60 °C. Results were analyzed using BioRad CFX Maestro 2.0 software (Bio-Rad Laboratories). The test was validated according to EPPO standard PM 7/98 (5) (EPPO, 2021) (Table S1).

To obtain (near) complete genome sequences Illumina sequencing was performed as described by Schoen et al. (2023). Total RNA extracts from leaf samples were sequenced on a NextSeq 500 or NovaSeq 6000 platform as 150 paired-end reads. Sequences of each genome segment were aligned with available (near) complete RNA1 and RNA2 sequences of TRSV from NCBI GenBank using MAFFT (Katoh et al., 2002) in Geneious Prime 2021.1.1 (Biomatters, New Zealand). External gaps and missing sequences were recoded as question marks.

Soil sampling for the detection of Xiphinema spp. was performed with an auger of 2 cm diameter and 40 cm depth. From each field an area of 1/3 ha was selected and a composite sample of 60 cores was taken in a systematic grid pattern, resulting in c. 7 L soil (EPPO, 2009). The soil samples were transported to the laboratory of NIVIP and kept at 4 °C until processing. From each thoroughly mixed composite soil sample a subsample of 200 ml was taken for nematode extraction by Oostenbrink elutriation with settings and sieves for extraction of larger nematodes (EPPO, 2013). The extracted suspension was microscopically examined for the presence of Xiphinema species, identification was performed based on morphological characteristics.

To gain insight in sequence similarity and clustering, phylogenetic analyses of TRSV isolates were performed with Bayesian inference (BI) using MrBayes v. 3.2.6 (Huelsenbeck & Ronquist, 2001). To determine the optimal model to construct the phylogenetic trees, for each genome segment a model test was performed in CLC genomic workbench (Qiagen, Denmark), using default settings. Subsequent analysis was performed with a random starting tree and four Monte Carlo Markov chains for 1.1 × 106 generations under the GTR model with G distribution. Trees were sampled every 200 generations. The first 5 × 104 generations were discarded as burn-in, and the remaining trees combined to generate a 50% majority rule consensus tree that represented posterior probabilities.

Table 3 gives an overview of the findings of TRSV in ornamentals in the Netherlands from 1997 until 2020 from surveys in The Netherlands and from imported plants. No virus-like symptoms were observed on any TRSV-infected plants, except for several Hemerocallis plants (2006) which, after sequencing in 2022, appeared to be co-infected by a novel putative Luteovirus (results not shown). It remains unclear if one or a combination of these viruses evoked the symptoms observed. Regarding the surveys for Xiphinema spp., the absence of X. americanum s.l. in the Netherlands was confirmed by analyzing 487 soil samples (Table 2). Therefore, based on the absence of X. americanum s.l., vegetative propagation is considered the main mode of spread of TRSV in the Netherlands.

Table 3 Overview of TRSV findings in the Netherlands between 1997–2020
Full size table

Phylogenetic analyses of all generated and available GenBank sequences show that isolates cluster different for their RNA1 and 2 sequences (Fig. 1), for example in the RNA1 tree MT563080 from Glycine max clusters with ON229972 from Hemerocallis and ON230016 from S. rosmarinus, while for RNA2 this sequence clusters with the various sequences from Cucumis melo. Although sequence clustering is different for RNA1 and 2, there are several groups where sequences cluster together in both trees, indicating that these isolates might have a common origin. For example, both RNA sequences from A. reptans cluster together with the sequences from Bacopa, Celosia, Lobelia, and Portulaca. Although Celosia was imported from Denmark, A. reptans, Bacopa, Lobelia, and Portulaca were imported from Israel, suggesting they may have a common origin. Only the A. reptans sequence from Guatemala clusters outside this group, suggesting a different origin of the isolate. Furthermore, as could be expected, sequences from C. melo from the same field in Texas cluster together (Tabara et al., 2021), as well as the sequences of Hemerocallis from 2017–2018. In contrast, the sequences from Hemerocallis from 2006–2007 do not cluster with those from 2017–2018, and sequences from I. germanica cluster in different clades which is more apparent for RNA1 than for RNA2. The clusters from I. germanica are not cultivar specific, i.e., sequences from the same cultivars group in different clusters and most clusters contain sequences of at least two cultivars. Overall, the similarity in sequences within the same host and year of finding, irrespectively of location, substantiates that TRSV was spread by vegetative propagation. Additionally, the fact that genotypes differ between different plant species indicates that TRSV had been able to enter the Netherlands on multiple occasions.

Fig. 1
figure 1

Bayesian inference tree of (near) complete TRSV sequences from this study (bold) and NCBI GenBank, using sequences of Aenoium ringspot virus as outgroup. Bayesian posterior probabilities (PP) are displayed at the nodes of the major braches. The scale bar indicates the number of substitutions per site. The trees Includes ON22940/41 NL Tulipa, identified in 1970 by the Laboratory of Flowerbulb research, kindly provided by Wageningen Plant Research. A) RNA1 B) RNA2

Full size image

The observed difference in clustering of RNA1 and 2 of TRSV is in line with Hily et al. (2021), who recently showed that the phylogenetic relationship between the Nepovirus subgroups was different when looking at either RNA1 or RNA2. While the division between subgroup A, B and C was clear when comparing RNA1 sequences, this was less clear for RNA2. RNA2 sequences of subgroup A, of which TRSV is a member, and C appeared scattered over different clades. Our results further substantiate this different clustering behavior of RNA1 and 2 and show the importance of using both RNA’s to determine phylogenetic relationships.

The results of the TRSV surveys revealed that the virus occurs without virus-like symptoms in various ornamental species. The phylogenetic analyses implicated that TRSV likely entered the Netherlands on multiple occasions and had further spread by vegetative propagation. This is supported by the absence of its nematode vector X. americanum s.l.. The previously mentioned reports of TRSV in ornamentals from several EU member states, suggest that this situation is likely similar in other countries as well. This is substantiated by the identification of TRSV in Celosia from Denmark and Ajuga and Coleus from Germany in this study. The approach of surveying regardless of symptoms might explain why findings in ornamentals are more frequent in the Netherlands compared to other EU countries. Therefore, systematic surveys in ornamentals regardless of symptoms are needed to determine the pest status in the EU, but since these have not been carried out by most other member states, the pest status is undetermined. Furthermore, the likely unnoticed spread by vegetative propagation in the EU underlines the importance of using virus-free material for vegetative propagation. Additionally, the TRSV reports in ornamentals raise the question whether the quarantine status in the EU is still applicable for TRSV. To qualify for a Q status, the pathogen needs to be absent in the EU, or if present, to have a restricted distribution, and have an unacceptable economic, social or environmental impact after introduction (Regulation (EU) 2016/2031; FAO, 2013). This study indicates that TRSV is probably more widespread in the EU than currently reported, at least in ornamental plants, and therefore does not meet the criteria of a quarantine pest. Moreover, the risks associated with the unnoticed presence in ornamentals seems limited, although an argument could be made that this is an unacceptable risk because the virus could spread by X. rivesi from these sources. Current legislation does not offer protection against the possibility of transmission by X. rivesi from ornamentals to the crops at risk, i.e., blueberry, grapevine and soybean. This implies that preventing the spread of TRSV requires large scale surveys in ornamentals and subsequent eradication measures in all EU countries. The question is whether it is desirable for EU phytosanitary agencies to put great effort into control of this virus. The number of outbreaks of quarantine pathogens has been increasing over the past years (European commission, 2022), while resources are limited. Therefore, another option which could be considered, is to regulate TRSV as a Regulated non-Quarantine pest (RNQP) (FAO, 2016) for blueberry, grapevine, and soybean. Moreover, in comparison to the current quarantine status, an RNQP status will provide better guarantees of the absence of TRSV as it requires testing of plants for planting of those crops at risk. In view of these considerations, the evaluation of the regulatory status of TRSV is highly recommended.