Development of a One-Step Multiplex Reverse Transcription-Quantitative Polymerase Chain Reaction Assay for the Simultaneous Detection of Four Lily Viruses
Article information
Abstract
Lilium spp. is an important cut flower in international and Korean markets due to its unique characteristics and fragrances. Millions of lily plants are cultivated, imported, and exported globally. Lily plants, propagated vegetatively through bulbs or scales to maintain the same genetic characteristics of lily flowers, require a longer growth period, increasing their risk of contact with infectious pathogens. Viral infections have been reported to cause both quantitative and qualitative damage to lily production yields worldwide. Common lily-infecting viruses include lily mottle virus (LMoV), lily symptomless virus (LSV), cucumber mosaic virus (CMV), and plantago asiatica mosaic virus (PlAMV). While numerous detection techniques have accurately confirmed the incidences of these viruses in lily plants, a one-step multiplex reverse transcription-quantitative polymerase chain reaction (RT-qPCR) has not yet been established. In this study, a one-step multiplex RT-qPCR was optimized to detect these four viruses in lily leaf samples collected in the Republic of Korea. Optimal primer concentrations were 10, 5, 5, and 2 µM for PlAMV, LMoV, CMV, and LSV detection, respectively, with optimal probe concentrations of 2, 1, 1, and 2 µM. A temperature of 57°C produced the optimal results among six tested annealing temperatures. The developed probe-based multiplex RT-qPCR demonstrated higher sensitivity compared to the conventional multiplex RT-PCR, detecting viral RNAs at as low as 3×10-4, 3×10-5, 3×10-5, and 3×10-5 ng for LMoV, PlAMV, CMV and LSV, respectively. This optimized assay proved sensitive, specific, and reproducible in distinguishing and simultaneously detecting four lily viruses effectively in 24 lily field samples.
Introduction
Lilies are among the most popular ornamental flower crops cultivated worldwide. The lily flower industry in the Republic of Korea has flourished since 1980, leading to a high demand for lilies. Consequently, lilies have become one of the largest and most valuable exported agricultural products, with an annual export value of 4 million US dollars in 2021 (Ministry of Agriculture Food and Rural Affairs, 2022). Korean farmers, however, have difficulty making profits from the production due to the high cost of lily bulbs, which are imported from foreign countries (Jang and Kim, 2016). As lily plants are typically vegetatively propagated through bulbs or scales to maintain the same genetic characteristics of lily flowers, the imported bulbs have proven unsuitable for cultivation beyond 3 years. This limitation arises from prevalent viral infections in the imported bulbs, leading them to be highly susceptible to virus infections. This has resulted in a decrease in lily cut flower production and a significant increase in the virus disease rates (Kim et al., 2019). Bulb propagation also requires longer time, leading to the increased rate of contact with infectious pathogens, making the maintenance of healthy lily plants extremely challenging (Van Tuyl et al., 2018). The four dominant viruses causing economic losses in lily production in Korea and worldwide are lily mottle virus (LMoV, Potyvirus), lily symptomless virus (LSV, Carlavirus), cucumber mosaic virus (CMV, Cucumovirus), and plantago asiatica mosaic virus (PlAMV, Potexvirus) (Asjes, 2000; Lawson, 2011; Lim et al., 2021; Xu and Ming, 2022). In the field, these viruses have been reported to infect either as a single or a mixed infection within the lily plants. In current scenarios, agriculturists often observe mixed infections, which constitutes approximately 70% of the viral diseases affecting lilies. The occurrence of single infections was comparatively lower, accounting for only 17.9% of the cases (Kim et al., 2019). Symptoms in the infected plants may differ from plant to plant depending on combination of viruses and seasonal changes, making identification or differentiation challenging for the naked eyes (Kwon et al., 2013). Mixed viral-infected lily plants typically exhibit severe symptoms, including severe leaf mosaic and dwarfing (Singh et al., 2008). Most importantly, there are no agrochemicals available to directly control viruses, unlike other plant pathogens (which can be addressed using fungicide or bacteriocide). To manage viral diseases, indirect methods, including insect viral vector control or the removal of diseased plants, are commonly applied (Jeong et al., 2014). Thus, the development of an accurate and effective virus diagnostic method is essential for the prompt detection and elimination of viral infections.
To detect lily viruses, a range of serological/immunological and molecular methods, including enzyme-linked immunosorbent assay (Kim et al., 2012), reverse transcription-polymerase chain reaction (RT-PCR) (Aravintharaj et al., 2017; Kwon et al., 2013; Niimi et al., 2003; Xu and Ming, 2022; Yoo and Jung, 2014), immunocapture-RT-PCR assay (Zhang et al., 2017), reverse transcription loop-mediated isothermal amplification (He et al., 2016; Zhang et al., 2020, 2022; Zhao et al., 2018), and quantitative PCR (qPCR or real-time PCR) (Lim et al., 2010; Xu et al., 2021) are widely used. Among these detection methods, only RT-PCR and qPCR have been reported to be able to detect more than three lily viruses simultaneously in one reaction. The drawback of RT-PCR is due to the need of agarose gel-based visualization, leading to a higher risk of post-contamination (Jeong et al., 2014; Lim et al., 2016). RT-qPCR is used for high throughput virus detection, rapid target-specific amplicon detection, and accurate quantification without the additional step of gel visualization. Although RT-qPCR has been universally applied to detect plant viruses, a one-step RT-qPCR assay has not been applied to detect viruses in lilies. To our knowl-edge, there is only one research paper using probe-based qPCR for multiplex lily viruses detection simultaneously (Xu et al., 2021). In comparison to qPCR, a one-step RT-qPCR is rapid, time-saving, labor-saving, and particularly cost-effective as this method does not require multiple steps. One-step RT-qPCR is also more reliable as it uses RNA samples directly as a template, similar to actual field samples (Thermo Fisher Scientific, 2016). On the other hand, RNA viruses contain RNA as their genetic material, using RT-qPCR is more preferable and accurate to diagnose than using qPCR. In this study, a probe-based one-step multiplex RT-qPCR was developed for precise lily viruses’ detection and its sensitivity was compared to that of the multiplex RT-PCR. Subsequently, the assay was applied to test on lily field samples.
Materials and Methods
RNA extraction.
Lily leaves in Sorbonne cultivar with viral symptoms, including vein clearing, leaf rolling, spotting, mosaic, and yellow streaking, were collected from Wanju, Republic of Korea. Total RNA extraction from the lily leaves was performed using the easyBLUETM Total RNA Extraction Kit (iNtRON Biotechnology, Seongnam, Korea) following the manufacturer's guidelines. RNA quality and RNA concentration were measured by using the Gen5TM Take3 Module program and Epoch microplate spectrophotometer (BioTek, Winooski, VT, USA). The lily leaf sample (S2), that was used for optimization assay, was confirmed to be infected with LMoV, PlAMV, CMV, and LSV by conventional RT-PCR and sequencing.
Primer and probe design.
Primer sets targeting the whole coat protein region of LMoV, PlAMV, CMV, and LSV were designed, and the PCR products were cloned and sent for sequencing. The sequencing results of LMoV, PlAMV, CMV, and LSV were deposited in NCBI library under accession numbers LC773423, LC773709, LC776735, and LC774317, respectively. These sequences were then aligned with other published sequences in National Center for Biotechnology Information (NCBI) library by using Clustal Omaga (Clustal Omega; EMBL-EBI, Hinxton, UK) as shown in Supplementary Figs. 1-4. Only the conserved regions were chosen to design RT-qPCR primers and probes. All RT-qPCR primers have been designed depending on alignment results without using any software programs as following the standard condition; the melting temperature between 50°C and 65°C, 40-60% of G + C content, 18-25 bases length, below 150 base pairs of amplicon size, and trying to avoid 4 identical nucleotide bases or excessive dinucleotide repeats. For probes, the length was designed in a range of 24-27 bases with a location that is close to the primer but does not overlap (Table 1).
Designed primers and probes of LMoV, PlAMV, CMV, and LSV for the multiplex RT-qPCR assay
One-step RT-qPCR assay optimization.
Probe-based RT-qPCR reactions were conducted using the CFX Opus 96 Real-Time PCR System (BioRad Laboratories, Hercules, CA, USA). The reaction consisted of 5 μl of TOPreal™ Multi-Probe RT-qPCR Kit (Enzynomics, Daejeon, Korea), 1 μl of each forward and reverse primer, 1 μl of probe, 2 μl of RNA template, and 1 μl of water, in a total volume of 20 μl. The reactions were performed under the following thermal conditions: comple-mentary DNA synthesis at 50°C for 30 min, pre-denaturation at 95°C for 10 min, and PCR cycle: 40 PCR cycles consisting of 5 sec at 95°C and 30 sec at 55°C with plate read. The optimization included various primer/probe concentrations (0.2 µM to 10 µM) and annealing temperatures (51°C, 53°C, 55°C, 57°C, 59°C, 61°C). Five points of the 10-fold serial dilution of RNA sample, ranging from 3 ng to 3×10-4 ng, were prepared and used for optimization.
Standard curve, reproducibility, and sensitivity assay.
To construct the standard curves of LMoV, PlAMV, CMV, and LSV to evaluate the efficiency and sensitivity of this multiplex RT-qPCR assay, 10-fold serial dilutions of the RNA sample ranging from 3 ng to 3×10-6 ng were prepared and used as templates. To ensure the reproducibility of the experiment, the inter-assay threshold cycles (Ct) values of the probe-based multiplex RT-qPCR were conducted. Data quantities of the samples were obtained by conducting the sample in triplicates on different days and across multiple experiments.
Validation of multiplex RT-qPCR assay.
To further validate the efficiency of the probe-based one-step multiplex RT-qPCR assay, a total of 24 field lily leave samples were tested. Different cultivar samples were tested including Sorbonne, Woori Tower, Siberia, and Camilla cultivars. These samples were collected from three different locations in South Korea, including Wanju (North Jeolla), Jeonju (North Jeolla), and a lily farmer's field in Gangneung (Kangwon). All of the samples were suspected to be infected with at least one of these four lily viruses (LMoV, PlAMV, CMV, and LSV) either in a single or mixed infection.
Results
Specificity of primer and probe.
As mixed infection of LMoV, PlAMV, CMV, and LSV usually happens on lily plants, the specificity of a primer set of each virus should be conducted. Plasmid DNA of each virus was used as template and positive control, while ultrapure water served as a negative control. Since there was no healthy lily samples, non-template control was not conducted. Each primer set of each virus was tested against its own positive sample and other viruses’ positive samples in separate reactions. As primers used in RT-qPCR assay were designed carefully based on the highly conserved region of each virus, the expected amplicon sizes were observed via the conventional PCR with 99 bp, 132 bp, 93 bp, and 92 bp for LMoV, PlAMV, CMV, and LSV, respectively. The primer designed for each virus showed the specific and singular band with the correct size for the corresponding virus, and the non-specific bands were not observed for other viruses (Fig. 1A). To ensure specificity, the designed primers and probes were retested using qPCR assay. As shown in Fig. 1B, only the positive samples of LMoV, PlAMV, CMV, and LSV showed the correct amplification curve. Ultrapure water, which served as a negative control, showed no amplification curves. All amplification curves with a Ct value greater than 36 were considered noise. These results indicated that the primer sets for each virus were specific to that virus and that no cross-contamination occurred in this reaction.
RT-qPCR primer specificity by conventional PCR and probe-based qPCR. A 100 bp ladder served as a marker to confirm the amplicon size for lily mottle virus (LMoV) with 99 bp (A 1), plantago asiatica mosaic virus (PlAMV) with 132 bp (A 2), cucumber mosaic virus (CMV) with 93 bp (A 3), and lily symptomless virus (LSV) with 92 bp (A 4). The reactions were run on 2% agarose gel electrophoresis at 100 volts for 40 min. As expected, only using the primer set corresponding to each specific virus showed a band, and no bands were observed in the other wells besides its own well (A). Similar to the conventional PCR, the probe and primer sets corresponding to each virus were tested with LMoV, PlAMV, CMV, and LSV samples, and the ultrapure water served as a negative control (2). LMoV primer with 99 bp (B 1), PlAMV primer with 132 bp (B 2), CMV primer with 93 bp (B 3), and LSV primer with 92 bp (B 4) showed the specific amplification curve with the earlier Ct value when tested with its own virus sample. No amplification curve was observed when each primer was tested with negative control (B 5). RT-PCR, reverse transcription-polymerase chain reaction; qPCR, quantitative polymerase chain reaction.
One-step RT-qPCR optimization.
Among the combinations of primer/probe concentration ranging from 0.2 µM to 10 µM, the final optimal concentration was P15. This combination consisted of 10, 5, 5, and 2 µM of primer (500, 250, 250, and 100 nM in the final concentration) and 2, 1, 1, and 2 µM of probe (100, 50, 50, and 100 nM in the final concentration) for multiplex RT-qPCR detection of PlAMV, LMoV, CMV, and LSV, respectively (Fig. 2). This combination gave better and more accurate Ct values for all of the viruses com pared to the other combinations shown in Supplementary Tables 1 and 2. The standard curve of each virus showed a high linearity, with the coefficient R2 of all viruses being 0.999 except for PlAMV, which was 0.997. This combination (P15) produced the most similar amount of relative fluorescence unit (RFU) for all viruses.
Optimization of the probe-based RT-qPCR primer/probe concentration. Various concentrations of primer and probe were tested with an annealing temperature at 55°C and five concentrations ranging from 3 ng to 3×10-4 ng of the serial dilution of RNA sample. All of the five concentrations of the RNA samples were detected by the P15 combination with a similar gap in the amplification curve. The coefficient R2 for all of the viruses was close to 1. LMoV, lily mottle virus; PlAMV, plantago asiatica mosaic virus; CMV, cucumber mosaic virus; LSV, lily symptomless virus; RT-qPCR, reverse transcription-quantitative polymerase chain reaction.
To optimize the annealing temperature, a probe-based multiplex RT-qPCR was performed at gradient temperatures between 51°C to 61°C with optimum primer/probe concentration (Fig. 3). The optimal annealing temperature was expected to provide the earliest Ct value with an acceptable RFU. The annealing temperature of 59°C gave the earliest Ct for LMoV, PlAMV, and LSV with values of 22.37, 21.34, and 19.35, respectively. At 59°C, CMV had a Ct value of 20.33, which was later than the annealing temperature of 57°C. The annealing temperature of 57°C gave the second highest Ct value with values of 22.64, 21.42, 20.06, and 19.59 for LMoV, PlAMV, CMV, and LSV, respectively. Although 59°C gave the earliest Ct value, the difference in the Ct value was only 0.3 at maximum from 57°C. In addition, annealing temperatures of 57°C produced the highest RFU for all viruses, except LSV, and was higher than annealing at 59°C. Thus, the annealing temperature of 57°C was chosen to further develop the probe-based multiplex RT-qPCR assay.
Annealing temperature optimization of the probe-based multiplex RT-qPCR. The assay was performed with gradient annealing temperatures, including 51°C, 53°C, 55°C, 57°C, 59°C, and 61°C. A 0.3 ng concentration of RNA sample and the optimum concentration of the primer and probe were used for testing. The different annealing temperatures resulted in different Ct values and relative fluorescence units (RFU) for lily mottle virus (LMoV) (1), plantago asiatica mosaic virus (PlAMV) (2), cucumber mosaic virus (CMV) (3), and lily symptomless virus (LSV) (4). While an annealing temperature of 59°C yielded the earliest Ct value for all viruses except CMV compared to 57°C, the difference in Ct values between these two temperatures was not substantial. Interestingly, an annealing temperature of 57°C resulted in the highest RFU for all viruses except LSV (depicted in the green graph). RT-qPCR, reverse transcription-quantitative polymerase chain reaction; Ct, threshold cycles.
Evaluation of the reproducibility of one-step probe-based multiplex RT-qPCR.
After optimization, the optimum concentration and annealing temperature were applied to conduct three different experiments at three different days for inter-assay. As shown in Table 2, the coefficient of variation (CV) of inter-assay for LMoV, PlAMV, CMV, and LSV were 1.448%, 1.739%, 0.661%, and 2.633% for the RNA concentration of 3 ng/µl, respectively. The CV of the 0.3 ng/µl concentration of the RNA sample was 2.481%, 2.618%, 3.473%, and 3.713% for LMoV, PlAMV, CMV, and LSV, respectively. The values of the CV remained within the range of the reaction reproducibility.
Inter-assay reproducibility of one-step probe-based multiplex RT-qPCR assay
Comparison of the sensitivity of the multiplex RT-PCR and the probe-based multiplex RT-qPCR.
To assess the sensitivity of each method, the RNA templates were serially diluted from 3 to 3×10-6 ng. According to Fig. 4A, the conventional multiplex RT-PCR could detect as low as 3×10-4 ng for LMoV, 3×10-3 ng for PlAMV, 3×10-3 ng for CMV, and 3×10-5 ng for LSV. The maximum Ct value of 36 was omitted for the one-step probe-based multiplex RT-qPCR. The minimum RNA concentrations that this assay could detect were as low as 3×10-4 ng for LMoV with a Ct value of 34.92, 3×10-5 ng for PlAMV with a Ct value of 35.91, 3×10-5 ng for CMV with a Ct value of 35.56, and 3×10-5 ng for LSV with a Ct value of 35.90 (Fig. 4B). Compared to conventional multiplex RT-PCR, probe-based multiplex RT-qPCR is 100 times more sensitive for PlAMV and CMV detection. Although the sensitivity of LSV detection of multiplex RT-PCR is similar to probe-based multiplex RT-qPCR, the band on 3×10-5 ng well in conventional multiplex RT-PCR assay is weak and difficult to distinguish (Fig. 4A).
Detection limit of the conventional multiplex RT-PCR and the probe-based multiplex RT-qPCR assay. The RNA sample was serially diluted from 3 ng to 3×10-6 ng and used as a template. Running on a 2% agarose gel electrophoresis for 70 minutes, the multiplex RT-PCR assay showed the detection limit at 3×10-4, 3×10-3, 3×10-4, and 3×10-5 ng for lily mottle virus (LMoV), plantago asiatica mosaic virus (PlAMV), cucumber mosaic virus (CMV), and lily symptomless virus (LSV), respectively (A). The detection limit of the probe-based multiplex RT-qPCR assay was 3×10-4, 3×10-5, 3×10-5, and 3×10-5 ng for LMoV, PlAMV, CMV, and LSV, respectively (B). RT-PCR, reverse transcription-polymerase chain reaction; qPCR, quantitative polymerase chain reaction.
Validation of the multiplex RT-qPCR assay.
A survey of four lily viruses in field samples collected from three different locations in South Korea (Wanju, Gangneung, and Jeonju) using the developed probe-based multiplex RT-qPCR showed that all field samples were infected (Table 3). All Sorbonne lily samples had mixed infections except S1. Five out of the eight samples collected from the Wanju area were co-infected with three viruses, including LMoV, PlAMV, and CMV (S4, S6, S7, S8, and S9). Similarly, all of the Woori Tower cultivars collected from the Gangneung area were infected. All of these samples, except KW2, were co-infected by two viruses, PlAMV and CMV (KW1, KW3, KW4, KW5, KW6, KW7, and KW8). Among the samples collected from the Jeonju area, there were two lily cultivars, Siberia and Camilla. Out of the seven samples of the Camilla lily cultivar, three were co-infected (C1, C2, and C3), and four were single-infected (C3, C4, C5, and C7). The only Siberia sample was co-infected by two viruses, PlAMV and CMV (Si1).
Diagnosis of LMoV, PlAMV, CMV, and LSV in lilies from three different locations in South Korea using the developed one-step probe-based multiplex RT-qPCR
In summary, six samples were single-infected, 14 samples were co-infected with two viruses, and five samples were co-infected with three viruses. CMV was the most dominant virus with an 83% infection rate, most of which were co-infection (17 out of the 19 samples). PlAMV followed with a 67% infection rate, all of which were mixed infections. LMoV showed a 46% infection rate, with more than half of the samples being mixed infections. LSV was not diagnosed in the lily field samples used in this study.
Discussion
Ct values exceeding 36 were excluded due to the amplification graphs producing minimal RFU, resulting in the graphs being barely visible, as shown in Fig. 3 and Supplementary Table 1.
As the primer and probe play a crucial role in the probe-based multiplex RT-qPCR, the specificity of the primer needed to be conducted. Only one band was observed in this study in the well corresponding to its target virus, with no non-specific bands appearing in the other virus wells (Fig. 1). This indicated the specificity of each primer set, as confirmed by conventional PCR. However, although the early amplification curve was generated specifically for each virus in the probe-based qPCR, there were some background noises. This problem was attributed to the primer-dimer formation (Thermo Fisher Scientific, 2016).
Besides the primer and probe concentration, another factor that plays an important role in the probe-based multiplex RT-qPCR assay is the annealing temperature. The expected optimum annealing temperature should give the earliest Ct value and the highest value of RFU. In this study, the optimal temperature ranged from 55°C to 59°C, while the Ct value increased when the temperature was above 61°C. This indicates that there is an exact range of temperatures that provides the best amplification in RT-qPCR assay. This similarity happened in the diagnosis of the three latent viruses of pome by using RT-qPCR assay. Beaver-Kanuya and Harper (2020) discovered that the detection of apple mosaic virus, apple chlorotic leaf spot virus, and apple stem pitting virus was effective with temperatures between 57-60°C, while temperatures above 62-63°C prevented the amplification. Additionally, Beaver-Kanuya et al. (2021) had the same results after utilizing an RT-qPCR assay for the detection of the three carlaviruses infecting hops, including American hop latent virus, hop latent virus, and hop mosaic virus (HMV).
According to the Minimum Information for Publication of Quantitative Real-Time PCR Experiments guidelines, the reaction efficiency should be in the range of 90-110%, which corresponds to the range of a slope of −3.58 and −3.10 (Bustin et al., 2009). In this study, the reaction efficiencies for LMoV, PlAMV, CMV, and LSV were 99.3%, 86.0%, 89.3%, and 81.8%, respectively (Fig. 4B). PlAMV and LSV fell outside the theoretical range of the reaction efficiency, which may have been caused by resource competition; and this is a common problem in multiplex assays (Thermo Fisher Scientific, 2016).
The CV of LMoV, PlAMV, CMV, and LSV detection using RNA as a template for inter-assay of this study was considered acceptable since Kralik and Ricchi (2017) has suggested that the limit of quantification or limit of detection in the molecular diagnosis of microorganisms was CV <25%. Although the CV for inter-assay was within an acceptable range, some values, notably around 4%, were considered slightly high (Table 2). This may cause to the fact that RNA viruses are naturally easier to break down, maintaining stable results was challenging (Minchin and Lodge, 2019). Using plasmid DNA as a temple is more stable compared to RNA, however, obtaining it requires more resources, including multiple procedures, time-consuming, resource-intensive, and labor-intensive. Despite these drawbacks, using RNA as a standard template is most similar to actual RNA samples as it does not require modifications before the experiment, unlike the case with plasmid DNA. On the other hand, in vitro transcribed RNA is widely used in one-step RT-qPCR (Bertolini et al., 2008; Kokane et al., 2021; Laiton-Donato et al., 2023; Ruiz-Ruiz et al., 2007). Unfortunately, this method can be inaccurate, considering that the synthesis of the in vitro transcribed RNA requires a long process and is therefore resource-consuming, with the potential for errors at any step (Beckert and Masquida, 2011). A small mistake in any procedure yields an in vitro synthesized RNA that is unreliable to use as a standard. Moreover, since the nature of RNA is unstable, maintaining its accuracy over time is challenging (Thermo Fisher Scientific, 2016). The choice of template was considered one of the main challenges.
In conclusion, the one-step RT-qPCR assay may be preferable over the qPCR for detecting RNA viruses, given that RNA viruses contain RNA as their genetic material. Optimizing the assay with a template form similar to the actual sample proved to be easier, more efficient, and more accurate than using a template from a different form. Using the one-step RT-qPCR assay helps minimize the risk of post-contamination during the plasmid DNA modification process, saving time, reducing costs, and reducing the required labor. This study showed the successful development and establishment of a one-step probe-based RT-qPCR assay that is reproducible, reliable, and well-suited for simultaneous detection of LMoV, PlAMV, CMV, and LSV in lilies.
Notes
Conflicts of Interest
No potential conflict of interest relevant to this article was reported.
Acknowledgments
This work was supported by National Institute of Agricultural Sciences, Rural Development Administration, Republic of Korea through Crop Viruses and Pests Response Industry Technology Development Program (or Project), funded by Rural Development Administration (RDA) (no. PJ014947022023).
Electronic Supplementary Material
Supplementary materials are available at Research in Plant Disease website (http://www.online-rpd.org/).