PCR amplicon next generation sequencing (NGS) analysis offers a broadly applicable and targeted approach to detect populations of both high- or low-frequency virus variants in one or more plant samples. In this study, amplicon NGS was used to explore the diversity of the tripartite genome virus, Prunus necrotic ringspot virus (PNRSV) from 53 PNRSV-infected trees using amplicons from conserved gene regions of each of PNRSV RNA1, RNA2 and RNA3. Sequencing of the amplicons from 53 PNRSV-infected trees revealed differing levels of polymorphism across the three different components of the PNRSV genome with a total number of 5040, 2083 and 5486 sequence variants observed for RNA1, RNA2 and RNA3 respectively. The RNA2 had the lowest diversity of sequences compared to RNA1 and RNA3, reflecting the lack of flexibility tolerated by the replicase gene that is encoded by this RNA component. Distinct PNRSV phylo-groups, consisting of closely related clusters of sequence variants, were observed in each of PNRSV RNA1, RNA2 and RNA3. Most plant samples had a single phylo-group for each RNA component. Haplotype network analysis showed that smaller clusters of PNRSV sequence variants were genetically connected to the largest sequence variant cluster within a phylo-group of each RNA component. Some plant samples had sequence variants occurring in multiple PNRSV phylo-groups in at least one of each RNA and these phylo-groups formed distinct clades that represent PNRSV genetic strains. Variants within the same phylo-group of each Prunus plant sample had ≥97% similarity and phylo-groups within a Prunus plant sample and between samples had less ≤97% similarity. Based on the analysis of diversity, a definition of a PNRSV genetic strain was proposed. The proposed definition was applied to determine the number of PNRSV genetic strains in each of the plant samples and the complexity in defining genetic strains in multipartite genome viruses was explored.

One hundred Prunus trees, including almond (P. dulcis), apricot (P. armeniaca), nectarine (P. persica var. nucipersica), peach (P. persica), plum (P. domestica), purple leaf plum (P. cerasifera) and sweet cherry (P. avium), were selected from growing regions Australia-wide and tested for the presence of 34 viruses and three viroids using species-specific reverse transcription-polymerase chain reaction (RT-PCR) or polymerase chain reaction (PCR) tests. In addition, the samples were tested using some virus family or genus-based RT-PCR tests. The following viruses were detected: Apple chlorotic leaf spot virus (ACLSV) (13/100), Apple mosaic virus (ApMV) (1/100), Cherry green ring mottle virus (CGRMV) (4/100), Cherry necrotic rusty mottle virus (CNRMV) (2/100), Cherry virus A (CVA) (14/100), Little cherry virus 2 (LChV2) (3/100), Plum bark necrosis stem pitting associated virus (PBNSPaV) (4/100), Prune dwarf virus (PDV) (3/100), Prunus necrotic ringspot virus (PNRSV) (52/100), Hop stunt viroid (HSVd) (9/100) and Peach latent mosaic viroid (PLMVd) (6/100). The results showed that PNRSV is widespread in Prunus trees in Australia. Metagenomic high-throughput sequencing (HTS) and bioinformatics analysis were used to characterise the genomes of some viruses that were detected by RT-PCR tests and Apricot latent virus (ApLV), Apricot vein clearing associated virus (AVCaV), Asian Prunus Virus 2 (APV2) and Nectarine stem pitting-associated virus (NSPaV) were also detected. This is the first report of ApLV, APV2, CGRMV, CNRNV, LChV1, LChV2, NSPaV and PBNSPaV occurring in Australia. It is also the first report of ASGV infecting Prunus species in Australia, although it is known to infect other plant species including pome fruit and citrus.

The distribution of Ilarvirus species populations amongst 61 Australian Prunus trees was determined by next generation sequencing (NGS) of amplicons generated using a genus-based generic RT-PCR targeting a conserved region of the Ilarvirus RNA2 component that encodes the RNA dependent RNA polymerase (RdRp) gene. Presence of Ilarvirus sequences in each positive sample was further validated by Sanger sequencing of cloned amplicons of regions of each of RNA1, RNA2 and/or RNA3 that were generated by species specific PCRs and by metagenomic NGS of the full genome of the ilarviruses that were detected. Prunus necrotic ringspot virus (PNRSV) was the most frequently detected Ilarvirus, occurring in 48 of the 61 Ilarvirus-positive trees and Prune dwarf virus (PDV) and Apple mosaic virus (ApMV) were detected in three trees and one tree respectively. American plum line pattern virus (APLPV) was detected in three trees and represents the first report of APLPV detection in Australia. Two novel and distinct groups of Ilarvirus-like RNA2 amplicon sequences were also identified in several trees by the generic amplicon NGS approach. The high read depth from the amplicon NGS of the generic PCR products allowed the detection of distinct RNA2 RdRp sequence variant populations of PNRSV, PDV, ApMV, APLPV and the two novel Ilarvirus-like sequences. Mixed infections of ilarviruses were also detected in seven Prunus trees. Sanger sequencing of specific RNA1, RNA2 and/or RNA3 genome segments of each virus and total nucleic acid metagenomics NGS confirmed the presence of PNRSV, PDV, ApMV and APLPV detected by RNA2 generic amplicon NGS. However, the two novel groups of Ilarvirus-like RNA2 amplicon sequences detected by the generic amplicon NGS could not be associated to the presence of sequence from RNA1 or RNA3 genome segments or full Ilarvirus genomes, and their origin is unclear. This work highlights the sensitivity of genus-specific amplicon NGS in detection of virus sequences and their distinct populations in multiple samples, and the need for a standardised approach to accurately determine what constitutes an active, viable virus infection after detection by molecular based methods.

Shiraz disease (SD) is an economically important virus-associated disease that can significantly reduce yield in sensitive grapevine varieties and has so far only been reported in South Africa and Australia. In this study, RT-PCR and metagenomic high-throughput sequencing was used to study the virome of symptomatic and asymptomatic grapevines within vineyards affected by SD and located in South Australia. Results showed that grapevine virus A (GVA) phylogroup II variants were strongly associated with SD symptoms in Shiraz grapevines that also had mixed infections of viruses including combinations of grapevine leafroll-associated virus 3 (GLRaV-3) and grapevine leafroll-associated virus 4 strains 5, 6 and 9 (GLRaV-4/5, GLRaV-4/6, GLRaV-4/9). GVA phylogroup III variants, on the other hand, were present in both symptomatic and asymptomatic grapevines, suggesting no or decreased virulence of these strains. Similarly, only GVA phylogroup I variants were found in heritage Shiraz grapevines affected by mild leafroll disease, along with GLRaV-1, suggesting this phylogroup may not be associated with SD.

Rapid and reliable detection tools are essential for disease surveillance and outbreak management, and genomic data is essential to determining pathogen origin and monitoring of transmission pathways. Low virus copy number and poor RNA quality can present challenges for genomic sequencing of plant viruses, but this can be overcome by enrichment of target nucleic acid. A targeted whole genome sequencing (TWG-Seq) approach for the detection of cucumber green mottle mosaic virus (CGMMV) has been developed where overlapping amplicons generated using two multiplex RT-PCR assays are then sequenced using the Oxford Nanopore MinION. Near complete coding region sequences were assembled with ≥100× coverage for infected leaf tissue dilution samples with RT-qPCR cycle quantification (Cq) values from 11.8 to 38 and in seed dilution samples with Cq values 13.8 to 27. Consensus sequences assembled using this approach showed greater than 99% nucleotide similarity when compared to genomes produced using metagenomic sequencing. CGMMV could be confidently detected in historical seed isolates with degraded RNA. Whilst limited access to, and costs associated with second-generation sequencing platforms can influence diagnostic outputs, the portable Nanopore technology offers an affordable high throughput sequencing alternative when combined with TWG-Seq for low copy or degraded samples.

Apple mosaic virus (ApMV) and prune dwarf virus (PDV) are amongst the most common viruses infecting Prunus species worldwide but their incidence and genetic diversity in Australia is not known. In a survey of 127 Prunus tree samples collected from five states in Australia, ApMV and PDV occurred in 4 (3%) and 13 (10%) of the trees respectively. High-throughput sequencing (HTS) of amplicons from partial conserved regions of RNA1, RNA2, and RNA3, encoding the methyltransferase (MT), RNA-dependent RNA polymerase (RdRp), and the coat protein (CP) genes respectively, of ApMV and PDV was used to determine the genetic diversity of the Australian isolates of each virus. Phylogenetic comparison of Australian ApMV and PDV amplicon HTS variants and full length genomes of both viruses with isolates occurring in other countries identified genetic strains of each virus occurring in Australia. A single Australian Prunus infecting ApMV genetic strain was identified as all ApMV isolates sequence variants formed a single phylogenetic group in each of RNA1, RNA2, and RNA3. Two Australian PDV genetic strains were identified based on the combination of observed phylogenetic groups in each of RNA1, RNA2, and RNA3 and one Prunus tree had both strains. The accuracy of amplicon sequence variants phylogenetic analysis based on segments of each virus RNA were confirmed by phylogenetic analysis of full length genome sequences of Australian ApMV and PDV isolates and all published ApMV and PDV genomes from other countries.