Comprehensive Virome Analysis of Commercial Lilies in South Korea by RT-PCR, High-Throughput Sequencing, and Phylogenetic Analyses

Abstract

Viral diseases pose a significant threat to lily (Lilium spp.) cultivation; however, large-scale assessments of virus prevalence and diversity in South Korea are limited. This study combined RT-PCR surveys, high-throughput sequencing (HTS), and analyses of 48 lily hybrid transcriptomes to characterize the lily virome. RT-PCR screening of 100 samples from 13 regions showed that 87% were infected, primarily with lily mottle virus (LMoV, 65%), Plantago asiatica mosaic virus (PlAMV, 34%), cucumber mosaic virus (CMV, 34%), and lily symptomless virus (LSV, 25%). Mixed infections were approximately twice as frequent as single infections and were associated with greater symptom severity, particularly in triple-virus combinations. High-throughput sequencing expanded detection to six viruses, including milk vetch dwarf virus (MDV) and lily virus B (LVB), the latter confirmed as a variant of strawberry latent ringspot virus (SLRSV). Near-complete genomes of several viruses were assembled and validated through RT-PCR. Transcriptome mining identified eight virus species across 26 cultivars; PlAMV was the most common, and viral loads varied significantly among hybrids. Phylogenetic analyses revealed close relationships between Korean and Chinese isolates and host-related clustering in PlAMV. These findings highlight the complexity of lily viromes in South Korea and provide essential resources for diagnostics, disease management, and biosecurity.

Keywords: RT-PCR; high-throughput sequencing; lilies; plant viruses; transcriptome analysis.

Figure 1

Virus detection in 100 lily samples from South Korea using RT-PCR. (A) Map of the 13 regions where lily leaves were collected for virus detection. (B) Agarose gel images of RT-PCR products detecting four viruses: CMV, LMoV, LSV, and PlAMV. M = DNA ladder, P = positive control, N = negative control, lanes 1–6 = individual lily samples. Complete RT-PCR results for virus detection are provided in Table S2.

Figure 2

Proportion of virus infections in lily samples and regions of South Korea. (A) Proportion of samples positive for the four major lily viruses. (B) Proportion of lily samples co-infected with multiple viruses. (C) Proportion of virus infections in each geographical region.

Figure 3

Representative symptoms of viral diseases in lilies caused by single or mixed virus infections. (A) Photographs showing diverse virus-induced symptoms under different infection combinations. Each image represents characteristic disease symptoms associated with specific viral infections. The labels containing non-English terms in the figure were used to identify the samples. (B) Symptom severity for different virus infections. Numbers indicate the symptom index: 0, symptomless; 1, mosaic; 2, necrosis; 3, necrotic spot; 4, necrotic spot and stripe; 5, necrosis, mosaic, and chlorosis. The different colors indicate the number of coinfected viruses.

Figure 4

Confirmation of HTS results by RT-PCR using virus-specific primers. (A) The lily actin gene was used as a positive control to verify total RNA quality. (B) RT-PCR using LMoV-specific primers showed amplification in samples 13, 15, and 17. (C) RT-PCR using PlAMV-specific primers showed amplification in samples 8, 11, 12, 13, 14, 17, and 18. (D) RT-PCR using MDV S segment-specific primers indicated amplification in the Medusa (Med) sample among different cultivars. (E) RT-PCR using LVB RNA1- and RNA2-specific primers showed amplification in two Yelloween plants.

Figure 5

Virus detection patterns in 48 lily hybrid transcriptomes. (A) Proportion of virus detection in 48 lily hybrid transcriptomes. (B) Box plots showing viral contig distribution for each virus. Each viral RNA segment is indicated by a different color. (C) Number of lily hybrid cultivars in which each virus was identified. Each viral RNA segment is indicated by a different color. (D) Number of virus species identified per lily hybrid cultivar.

Figure 6

Proportion and quantification of viruses in lily transcriptome data. (A) Proportion of viral reads compared to total transcriptome reads. (B) Pie chart showing the relative abundance of individual identified viruses by FPKM. (C) FPKM values of viruses in each lily transcriptome. The red and blue bars indicate the highest and lowest values. (D) Viral composition per lily transcriptome.

Figure 7

Phylogenetic relationships of LMoV, LAV1, and LSV genomes obtained in this study compared with known viral genomes. (A) Phylogenetic tree of 22 LMoV isolates, including six genomes generated in this study. The tree was constructed using the best-fit substitution model identified by the Bayesian information criterion (BIC): GTR + F + I + G4. (B) Phylogenetic tree of four LAV1 genomes, including one newly sequenced genome from this study. The best-fit model selected by BIC was F81 + F. (C) Phylogenetic tree of 27 LSV genomes, including five newly identified genomes from this study. The SYM + G4 model was determined as the best fit by BIC for tree construction. Each group is indicated by a different color. The red box represents the viral genome derived from this study.

Figure 8

Phylogenetic relationships of PlAMV genomes obtained in this study compared with known isolates. Phylogenetic tree of 52 PlAMV isolates, including nine new genomes identified in this study. The tree was constructed using the best-fit substitution model determined by Bayesian information criterion (BIC): GTR + F + I + G4. Each group is indicated by a different color. The red box represents the viral genome derived from this study.

Figure 9

Phylogenetic relationships of SLRSV genomes obtained in this study compared with known isolates. (A) Phylogenetic tree of 30 SLRSV RNA1 isolates, including one novel genome identified in this study. The tree was constructed using the best-fit substitution model identified by the Bayesian information criterion (BIC): GTR + F + I + G4. (B) Phylogenetic tree of eight SLRSV RNA2 isolates, including one newly identified genome from this study. The best-fit substitution model determined by BIC was TIM2 + F + I. Each group is indicated by a different color. The red box represents the viral genome derived from this study.

Figure 10

Phylogenetic relationships of CNSV genomes obtained in this study compared with known isolates. (A) Phylogenetic tree of 25 CNSV RNA1 isolates, including two novel genomes identified in this study. The tree was constructed using the best-fit substitution model identified by the Bayesian information criterion (BIC): TIM2 + F+I + G4. (B) Phylogenetic tree of 31 CNSV RNA2 isolates, including three newly identified genomes from this study. The best-fit substitution model determined by BIC was TVM + F + I + G4. Each group is indicated by a different color. The red box represents the viral genome derived from this study.