Introduction

Narcissus (commonly known as narcissus, daffodils, and jonquil), belongs to the family Amaryllidaceae, and includes some of the most popular ornamental flowers in the world. Narcissus are vegetatively propagated ornamental bulbous plant with attractive flowers and pleasant scents. The genus Narcissus comprises 63 species, including many natural hybrids that are vegetatively propagated (Sochacki and Chojnowska 2016). In Korea, Narcissus are commonly grown in public gardens and flowerbeds, in addition to being used as cut flowers and potted plants. Narcissus are mainly imported from the Netherlands (https://okminwon.pqis.go.kr/ minwon/information/statistics.html), a major producer of bulbs plants. Narcissus plants propagated using bulbs are likely to harbor diverse viruses and are vulnerable to viral infections that can lead to economic losses (Wylie and Jones 2012). In this study, the presence of viruses was investigated using the bulb scale of Narcissus that are traded internationally and available on the Korean market.

Several genera including Carlavirus, Macluravirus, Nepovirus, Potexvirus, and Potyvirus have been reported infecting Narcissus (Probowati et al. 2022). Many viruses belonging to the genus Potyvirus that infect Narcissus plants have also been reported. They are cyrtanthus elatus virus A (CyEVA) (Kumar et al. 2015), hippeastrum mosaic virus (Ágoston et al. 2020), lily mottle virus (Liu et al. 2008), narcissus degeneration virus (NDV) (Chen et al. 2007), narcissus late season yellows virus (NLSYV) (Mowat and Dawson 1987), narcissus yellow stripe virus (NYSV) (Chen et al. 2006), and ornithogalum mosaic virus (OrMV) (Zheng et al. 2009). The most prevalent Narcissus-infecting potyviruses are CyEVA, which causes severe chlorotic leaf stripe symptoms; NDV, which causes streak mosaic symptoms; NLSYV, which causes narrow chlorotic streaks and elliptical markings on the leaves; NYSV, which causes chlorotic stripes on the leaves and flower stalks and reduced bulb size; and OrMV, which causes streaks, yellow stripes, and tip necrosis on leaves (Brunt 1966;Ohshima et al. 2016;Raj et al. 2020). A study using virome analysis revealed that eight different potyviruses out of 13 viruses, including three potexviruses and one macluvirus, were detected when ornamental bulb plants such as daffodils or tulips were tested (Wylie et al. 2019). As Narcissus are likely to harbor co-infections of potyviruses (Ohshima et al. 2016;Probowati et al. 2022), we investigated which virus species infecting narcissus are internationally traded and imported into the Korean market. Narcissus-infecting viruses reported in Korea have not been frequently addressed, except for the tobacco rattle tobravirus detected in Narcissus, Gladiolus, and Crocus (Shin et al. 2002). However, Narcissus-infecting potyviruses have been found in diverse Amaryllidaceae species, including Narcissus, in many countries, including China, Europe, Japan, India, and New Zealand (Chen et al. 2003;Probowati et al. 2022;Wylie et al. 2019). As international trade in ornamental bulbs has increased, healthy bulb selection and planting are important for healthy domestic bulb cultivation. We aimed to screen potyviruses as a feasible contagion in Narcissus bulbs and to acquire a practical technique for the detection of Narcissus-infecting potyviruses.

Materials and Methods

Bulb sources of Narcissus

In this study, 38 Narcissus bulb samples purchased from a domestic market in Korea were used to detect potyviruses. These bulbs were propagated and imported from the Netherlands, and the cultivars used were as follows: ‘Tete-a-tete,’ ‘Tete Boucle,’ ‘Salome,’ ‘Sound,’ ‘Dick Wilden,’ ‘Double Fashion,’ ‘Tahiti,’ and ‘Orangery.’ The size of the bulbs varied from 30 - 60 mm in diameter and comprised of differently categorized cultivars. The cultivars ‘Tete-a-tete’ and ‘Tete Boucle’ are N. cyclameneus daffodils; ‘Salome’ and ‘Sound’ are Small-cupped daffodils; ‘Dick Wilden,’ ‘Double Fashion,’ and ‘Tahiti’ are Double daffodils; and ‘Orangery’ is a Split corona daffodil, according to the classification of the Royal Horticultural Society (Hanks 2002).

Extraction of total RNAs from the bulb tissues

For total RNA extraction, fresh bulb tissue (150 mg) was ground in liquid nitrogen. Homogenized bulb tissue was treated with 1 mL of fruit-mate (TaKaRa, Shiga, Japan) for RNA purification, and the supernatant was collected after centrifugation. Thereafter, 200 μL of RNA extraction buffer (50 mM Tris-Cl pH 8.0, 0.1% SDS, 10 mM EDTA pH 8.0, 1% β-mercaptoethanol) and 200 μL of phenol : chloroform : isoamylalcohol (PCI, 25:24:1, v/v/v) were added, and centrifuged at 4°C at 12,000 rpm for 15 min. The supernatant was transferred to a fresh tube and an equal volume of 100% ethanol was added. After centrifugation, the RNA pellets were washed by 70% ethanol. Then, the air-dried pellet was resolved in 50 μL of sterilized distilled water. The quality and quantity were checked using a Nanodrop spectrophotometer (Thermo Fisher Scientific, Waltham, Massachusetts, USA).

Reverse transcription (RT)

First-strand cDNA was synthesized at 42°C for 60 min in a reaction mixture (20 μL) containing 5× reaction buffer (250 mM Tris-HCl pH 8.3, 250 mM KCl, 20 mM MgCl₂, 50 mM DTT), 1 μL of 2.5 mM dNTP mix (TaKaRa, Shiga, Japan), 0.5 μL of 400 μg/μL random hexamer (Qiagen, Hilden, Germany), 4 μL of total RNA (5 μg), and 100 U of RevertAid reverse transcriptase (Thermo Fisher Scientific, Waltham, Massachusetts, USA).

Polymerase chain reaction (PCR) using universal primers for potyvirus

To investigate the presence of a feasible virus in Narcissus bulbs, PCR was performed using poty-universal primers (Zheng et al. 2010) (Table 1) in a reaction mixture (50 μL) containing 10× PCR buffer (100 mM Tris-HCl pH 8.3 at 25℃, 500 mM KCl, 15 mM MgCl₂), 1 μL of 2.5 mM dNTP mix, 1 μL of 10 pmol/μL poty-universal primer set, and 5 μL of cDNA and 1 U of Taq polymerase (TaKaRa, Shiga, Japan). PCR was performed in a thermal cycler (T100™ Thermal Cycler, Bio-Rad Laboratories, Hercules, California, USA) under the following conditions: 3 min at 95°C, then 30 cycles at 95°C for 30 s, 45°C for 30 s, 72°C for 60 s, and finally 5 min at 72°C. After the PCR reaction, 3 μL of PCR products were separated on a 1.0% agarose gel at 135 V for 18 min. The gel was stained with ethidium bromide (EtBr) and visualized using a gel doc system (Gel Doc XR+ System, Bio-Rad Laboratories, Hercules, California, USA). A 100 bp DNA ladder (Thermo Fisher Scientific, Waltham, Massachusetts, USA) was used as a reference.

Sequencing of PCR products amplified using poty-universal primers

PCR fragments were ligated with the pGEM-T Easy vector using T4 DNA ligase (Promega, Madison, Wisconsin, USA). The ligation mixture was transformed into Escherichia coli strain DH5α (RBC Bioscience, New Taipei City, Taiwan). After transformation, the recombinant plasmids were selected and prepared using the alkaline lysis method (Sambrook et al. 1989). The selected plasmid DNA sequences were determined by standard sequencing at Macrogen, Inc. (Seoul, Korea) and analyzed using the BLAST program from the National Center for Biotechnology Information (NCBI).

Design of primers for specific virus detection and PCR

According to the nucleotide sequences generated from the PCR products using poty-universal primers, sequence homology indicated that three different potyviruses (CyEVA, NLSYV, and NYSV) were present in the samples. Based on these findings, primers for these three potyviruses were designed based on the sequences of the corresponding viral genomes from the NCBI database. Sequence alignment was performed using the DNAMAN software package (version 5.1; Lynnon Biosoft, Quebec, Canada). Designated target regions of CyEVA, NLSYV, and NYSV were selected for complete coat protein gene sequencing (Table 1). The specificity of the designed primer sets was verified using NCBI primer-BLAST. RT-PCR was executed under the same conditions as above, except for the annealing temperature, owing to the replacement of the virus-specific primers (Table 1). After visualization of the amplicons under UV light, subsequent cloning of the PCR product and sequencing were performed, as described above.

Sequences generated from Narcissus isolates of CyEVA, NYSLV, and NYSV, and phylogenetic construction

Nucleotide sequences of the viruses obtained in this study were compared with other virus isolates in the NCBI database using nBLAST. All isolates used for sequence comparison were shown in Fig. 3, including the viruses identified in this study. Nucleotide sequences from all isolates used for sequence alignment were trimmed for comparison with those obtained in this study. Nucleotide sequences were translated into amino acid (aa) sequences, and the aa sequences of the complete CP genes of CyEVA, NLSYV and NYSV were used to establish a distance matrix for analysis after sequence alignment. The sequences were aligned using ClustalW in MEGA software (version 10.2; PA, USA). Pairwise alignment was performed with a gap-opening penalty of 10, a gap extension penalty of 0.1 and multiple alignment was performed with a gap-opening penalty of 10, a gap extension penalty of 0.2 in an amino acid analysis and with a gap-opening penalty of 15, a gap extension penalty of 6.66 in a nucleotide analysis. Pairwise distance matrix was generated using the poisson model in an amino acid analysis. Phylogenetic tree was constructed by the neighbor-joining method using MEGA software. The dataset was subjected to 1,000 bootstrap replicates.

Results

Screening and identification of potyviruses in bulb tissues

RT-PCR was performed to investigate the presence of potyviruses. PCR fragments of 350 bp of the nuclear inclusion protein-coding sequence were amplified from samples using poty-universal primers (Table 1). In Narcissus, specific fragments using poty-universal primers were amplified in 17 of the 38 samples (‘Tete-a-tete’ 1−4, ‘Tete Boucle’ 1−5, ‘Salome’ 1 and 2, ‘Dick Wilden’ 3, ‘Tahiti’ 1−3, and ‘Orangery’ 1 and 3) (Fig. 1). The PCR fragments were cloned, sequenced, and analyzed. BLAST analysis showed that the tentative potyvirus sequences from the Narcissus cv. Orangery 1 shared 92.4% nucleotide sequence homology with the CyEVA sequence (JQ723475), the tentative potyvirus sequence of the Narcissus cv. Tahiti 1 shared 96.1% homology with the NYSV sequence (LC314396), and a tentative potyvirus sequence of the Narcissus cv. Salome 1 isolate shared 97.6% homology with the NLSYV sequence (MH886515) (Table 2). Based on these findings, each virus was subjected to subsequent analyses.

Application of virus-specific primers for diagnosis

Based on the nucleotide sequences obtained from the PCR fragments generated by poty-universal primers in Narcissus, virus-specific primers were designed and RT-PCR was performed to detect the presence of CyEVA, NLSYV, and NYSV in bulb samples. The primers were used to amplify the complete CP gene sequences of CyEVA (790 bp), NLSYV (854 bp), and NYSV (834 bp) (Table 1). In Narcissus, CyEVA was only detected in the ‘Orangery’ cultivar, and NLSYV was detected in 14 samples of five cultivars, i.e., ‘Tete-a-tete,’ ‘Tete Boucle,’ ‘Salome,’ ‘Dick Wilden,’ and ‘Tahiti.’ NYSV was detected in 12 samples of 3 cultivars, i.e., ‘Tete-a-tete,’ ‘Tete Boucle,’ and ‘Tahiti’ (Fig. 2). Co-infection with at least two potyviruses, NLSYV and NYSV, was also found to exist in 11 out of 38 samples in this study.

Identification of virus isolates obtained from Narcissus spp. based on phylogeny

To determine and estimate the genetic relationships between the viral isolates obtained in this study and other previously reported viral isolates, phylogenetic trees were generated from the distance matrix constructed from aa sequence alignments. The phylogeny among the CyEVA isolates and other potyviruses was generated (Fig. 3A). The genetic distance between the CyEVA isolates (Orangery1, 2) analyzed in this study and other potyviruses ranged from 0.504 to 0.741, which was greater than that between the two CyEVA (Orangery1, 2) and other CyEVA isolates (0.031−0.072) (distance matrix data not shown). This result indicated that the two CyEVA isolates analyzed here and other CyEVA isolates were more closely related than other potyvirus species. Based on the phylogenetic tree (Fig. 3A), the CyEVA isolates obtained from Narcissus in this study grouped with CyEVA isolates derived from diverse geographical origins (Australia, India, and Japan). Other potyviruses were distinctly differentiated from the CyEVA isolates. Therefore, the CyEVA isolates obtained from Narcissus in this study could be classified as CyEVA based on phylogenetic analysis.

Among the NLSYV isolates analyzed, a phylogeny was generated according to the calculated distance matrix (Fig. 3B). The genetic distance between the NLSYV isolates obtained in this study and other potyviruses ranged from 0.238 to 0.587, which was greater than that between the five NLSYV isolates analyzed here and other NLSYV isolates (below 0.080). In the phylogenetic tree, NLSYV isolates from diverse regions were clustered together (Fig. 3B). Other potyviruses were distinctively differentiated from the NLSYV isolate cluster; therefore, the five NLSYV isolates obtained from Narcissus in this study could be classified as NLSYV based on the phylogenetic analysis.

Regarding NYSV, the isolates (Tahiti1 and 3, Tete-a-tete1 and 4, Tete Boucle2 and 3, and Salome1) obtained in this study and other NYSV isolates from diverse origins were used to generate the phylogeny (Fig. 3C). Similar to CyEVA and NLSYV, the aa sequence distance matrix was calculated for the NYSV phylogeny. The genetic distance between the NYSV isolates analyzed here and other potyviruses ranged from 0.238 to 0.709, which was relatively greater than that between the NYSV isolates analyzed here and other NYSV isolates (0.011−0.207). In the phylogenetic tree, the NYSV isolates analyzed in this study were grouped with other NYSV isolates from Narcissus in China, India, and Japan. NYSV isolates were distinctly differentiated from the other potyviruses used in this comparison (Fig. 3C). Therefore, NYSV isolates obtained from Narcissus in this study were presumed to be NYSV based on our phylogenetic analysis.

Discussion

In this study, Narcissus samples comprising eight cultivars purchased randomly from a commercial market were investigated using RT-PCR to detect potyviruses in bulb tissues. Generally, Narcissus harbored diverse viruses in the bulb. Among the feasible viruses known to possibly infect Narcissus species, potyviruses were screened in the bulbs. A previous study based on virome analysis showed that 37 street bulb flowers were generally infected with potyviruses, such as NLSYV (Wylie et al. 2019). Raj et al. (2020) detected three different potyviruses, CyEVA, NYSV, and NDV, in N. tazetta leaves using potyvirus degenerate primers. Therefore, in this study, poty-universal primers were applied for potyvirus detection and individual PCR products were sequenced to identify the potyviruses.

NLSYV was the most ubiquitous potyvirus in our study, NYSV prevalence appeared to be slightly lower, whereas CyEVA was uncommon. As a member of the Potyvirus family, CyEVA was first reported by Inouye and Hakkaart (1980) in Vallota speciosa plants (Amaryllidaceae) in New Zealand. In the present study, CyEVA was detected in cv. Orangery, and sequences of CyEVA obtained here showed the highest sequence identities at the amino acid level with those reported from Australia (NC_017977) (Fig. 3A). Narcissus was originally thought to be the only host of NLSYV, but Clivia miniata (Lin et al. 2012;Wylie et al. 2010) and Sternbergia lutea (Ágoston et al. 2021) (both Amaryllidaceae) have been regarded as natural hosts. In this study, NYSV was frequently detected in various samples, and mixed infections with NYSV in cv. Tahiti, Tete-a-tete, and Tete Boucle were also observed (Fig. 2A and 2C).

According to a previous study, co-infection with NLSYV and NYSV was observed (12.5%) in Narcissus leaf samples (Probowati et al. 2022). Similarly, NLSYV and NYSV co-infection occurred in 11 out of 38 samples (28.9 %) in our study. We confirmed that CyEVA, NLSYV, and NYSV singly or doubly infected bulb samples when eight cultivars of Narcissus were tested using RT-PCR. In Indian Narcissus (N. tazetta cv. Paperwhite), three potyviruses, CyEVA, NYSV, and NDV, were identified using RT-PCR (Raj et al. 2020). Thus, complex infections among different types of potyviruses present in bulb organs appear to be a frequent phenomenon in various species of Narcissus. In future studies, co-infection may need to be investigated at the virome level using next-generation sequencing to investigate the interaction between ornamental bulbs in Amaryllidaceae, such as narcissus and amaryllis, and potyviruses. Additionally, eight cultivars in four different divisions (described in Materials and Methods) were used to determine whether a certain specificity exists between cultivar and virus type, however, a particular relationship between them was not clear, except for the case of CyEVA and Orangery cultivar. This might indicate that the viruses present in Narcissus bulbs are affected by the geographical region or country where bulbs are produced, rather than host parameters, such as species or cultivar. Since bulbs are often imported from abroad, it is probable that other exotic viruses are present. A more definitive conclusion would require extensive examination with a larger sample size, particularly using bulbs from the country of origin of imported plants. However, this result shows that it is important to support the domestic cultivation of healthy bulbs through adequate screening for viral infections in ornamental bulbs that are currently imported. The information generated in this study could be utilized to determine the type of potyviruses present in daffodil bulbs and could help accurately identify viruses such as CyEVA, NLSYV, and NYSV during cultivation, propagation, and quarantine using simple RT-PCR-based detection tools.