4. Discussion
The number of Prunus trees tested in this study was not extensive, however, the approach of using a combination of virus species-specific and genus/family-based RT-PCR tests, RT-PCR, or nested RT-PCR/PCR and HTS allowed us to test Australian grown trees for many Prunus infecting viruses and to examine their genetic diversity. This study showed that viruses not previously reported in Australia were present in Prunus species, many of which should be excluded from programs that produce high-quality planting material for the Australian stone fruit and almond industries. This study highlighted the importance of understanding virus genetic diversity to ensure diagnostic tests can detect all known genetic strains of a virus species. It also highlighted the value of metagenomic HTS as a single diagnostic test that can indiscriminately detect all viruses in a single sample. Australia is a large continent and Prunus trees are grown in diverse regions throughout the country, with cool-temperate to Mediterranean climates, sometimes 1000’s of kilometres apart. Some virus species occur in many regions and sometimes closely related virus strains appear in different and distant regions and dissemination of the viruses by propagation material is probably the primary pathway. Prunus material is imported from many countries including Europe and the USA and is also a likely contributor to the diversity of species and genetic diversity within species observed in this study.
One or a combination of viruses were found in 70% of the trees that were tested. PNRSV, PDV, ApMV and ACLSV were detected in this study. These viruses have long been known to occur in many
Prunus species in Australia and are included in certification testing for production of high health planting material, particularly in almonds [
50,
51]. Other endemic
Prunus viruses and viroids including APCLSV, APLPV, ASGV, CVA, HSVd and PLMVd, were mainly found in the stone fruit and ornamental
Prunus trees, but not in almonds. During the survey ApLV, APV2, CGRMV, CNRMV, LChV1, LChV2, NSPaV and PBNSPaV were detected in stone fruit species and this is the first report of these viruses in Australia. PNRSV was the most commonly detected virus, occurring in 52% of the samples tested. All other viruses that were detected occurred much less frequently (≤14%).
Except for LChV2, vectors are not known for most of the
Prunus infecting viruses that were found in this study. The known mealybug vectors of LChV2 (
Phenacoccus aceris and
Pseudococcus maritimus) [
52] are not known to occur in Australia. However, all viruses are transmitted vegetatively, during propagation and grafting, and some are transmitted in pollen, including PNRSV and PDV [
53] and these are likely to be the primary means of distribution of
Prunus infecting viruses in Australia. The high prevalence of PNRSV is likely to be due to a combination of the use of untested planting material, especially as the use of certified material is voluntary, and field transmission via pollen and thrips which occurs in Australia [
54,
55]. Although PDV is also pollen and vegetatively transmitted, similar to PNRSV, its low incidence (3%) in this study similar to other studies [
34] may be due to its
Prunus limited host range compared to PNRSV, which can also naturally infect other plant hosts [
56,
57,
58].
ASPV and CMV were not detected although they are known to occur in other crops in Australia. This result was anticipated as they appear to be minor viruses of
Prunus that are reported infrequently [
59,
60,
61] and do not seem to be tested in overseas certification programs. The following exotic viruses and viroids were not detected: ASSVd, ArMV, APV1, APV3, CLRV, CMLV, CRLV, CRMV, CTLaV, HSVd, PLMVd, PcMV, PRMV, PPV, RpRSV, SLRSV, TRSV, TBRV, TBSV and ToRSV. As a result, imported germplasm is likely to continue to be tested for these viruses during Australian PEQ. However, it is possible the viruses that were not detected are present in Australia but at low prevalence and too few trees were tested in this study to allow detection.
Although ASGV is known to infect
Prunus species in other countries [
62], this study presents the first evidence that ASGV, which frequently infects Australian pome fruit species [
63], can also infect
Prunus species (plum) in Australia. Infection could have been through imported infected planting material, but it may also have occurred through mechanical transmission through the sharing of equipment between
Prunus and other fruit crops such as citrus or pome fruit, where their growing regions overlap in Australia [
64].
It is difficult to estimate when ApLV, APV2, CGRMV, CNRMV, LChV1, LChV2, NSPaV and PBNSPaV, which were initially considered exotic, were introduced into Australia, but it is possible they have been present for many years but gone undetected until this study. Indeed, the Victorian plum tree infected with PBNSPaV, ACLSV, ASGV and PNRSV was estimated to be more than 60 years old, which suggests that PBNSPaV may have been present for more than 60 years. These viruses were most likely introduced into Australia before they were known and included as part of the Australian import molecular testing regime. The associated symptoms were probably not observed by biological indexing during Australian PEQ because symptoms were not well described for biological indicator species or the indicators were symptomless. This is especially the case for recently described viruses such as NSPaV, that are only now being discovered through the development and implementation of HTS in association with diseases that previously had an unknown aetiology [
24,
65]. Therefore, as new pathogens and updates in pathogen biology and disease aetiology occur, it is important that there is continual surveillance of the literature to ensure these PEQ pathogen lists remain current to support the Australian
Prunus industry’s biosecurity continuum.
The lack of detection of some viruses and viroids and low prevalence of others in Australia, especially those that are primarily spread through vegetative propagation, is likely to be a result of certification programs established in the 1970s [
51,
66] in combination with strict quarantine measures that adapt to include new virus species as they are reported. Testing is done during PEQ for many exotic viruses and exotic diseases with a suspected virus aetiology, using specific tests and/or biological indexing. In line with the remit of Australian quarantine, testing for non-regulated pests may not be done unless requested by importers or if a general test detects them. Detection of an endemic virus during PEQ may not preclude the entry of new germplasm but allows importers to decide how they would like to proceed with importation, directing it to a germplasm collection, undertaking virus eradication prior to distribution of material or discarding the material. New germplasm may enter certified pathogen-tested collections, which aim to ensure that the material distributed to industry is free of the significant viruses that occur in Australia like ACLSV, ApMV, PNRSV and PDV. The Australian almond industry tests almond trees and rootstocks annually for significant viruses, which minimises their spread and impact on production. However other Australian
Prunus germplasm collections may only be tested for endemic viruses upon entry into their collection or tested infrequently, increasing the risk of distributing virus-infected material to nursery and production industries, especially for those viruses like PNRSV, which are also spread by pollen [
67,
68].
Like many other studies [
69,
70,
71,
72], mixed infections of two or more viruses were also detected in 24/100
Prunus trees tested. PNRSV occurred in all mixed infections which is not surprising given that PNRSV is easily transmitted by pollen and vegetatively, and was the most frequently detected virus [
67,
68]. Interestingly, no mixed infections were detected in almond trees which may be attributed to the rigorous certification testing regime undertaken by the Australian almond industry. In contrast, mixed infections were frequently detected in stone fruits with most mixed infections, sometimes as many as five viruses were detected in cherry trees e.g in
Table 4 isolate TAS12, LV27, LV35 and LVV. Unlike the Australian almond industry, no formal certification virus testing program exists for other Australian
Prunus industries and trees may not be tested or are tested infrequently instead of annually for viruses. Until this survey was completed in 2015, most Australian
Prunus trees were tested only for ApMV, ACLSV, PDV and PNRSV because other viruses were not known to occur or were considered less significant. This infrequent testing, and only for a few viruses, especially before 2015, explains the high incidence of mixed infections observed in stone fruits in this study, especially in cherries, and highlights the need for these
Prunus industries to undertake a more proactive virus testing program in which testing is undertaken for all viruses now known to occur in Australia.
In this study, the reliability of RT-PCR and PCR diagnostic tests was assessed to ensure they were fit for purpose for routine diagnostic testing of Australian
Prunus trees for viruses. RT-PCR and PCR tests were chosen for this survey by comparing primers against available sequences in GenBank to determine their likelihood for detecting targeted virus, however, this process is only as good as the virus sequences available for comparison. Genetic similarity or variability of viruses and co-extracted nucleic acids may influence the reliability of RT-PCR and PCR tests resulting in false-positive or negative results. False-positive results can occur when the PCR primers have a high degree of similarity to other genetic material, including that of the host plant or other organisms within a geographic region [
73]. For instance, there was a large discrepancy in the detection of the trichoviruses APCLSV and ACLSV using the specific and genus/family-based RT-PCR tests, where all APCLSV and 6/13 ACLSV detections were only by the genus/family-based PDO RT-PCR test [
20]. Their amplicons had less than an 85% nucleotide identity with published APCLSV and ACLSV isolates, indicating that they were quite divergent. Similar variation in the detection of these trichoviruses has been previously reported and can be attributed to their high molecular variability which impacts the binding capacity of highly specific primers of the species-based RT-PCR tests, especially those used in this study [
74,
75,
76]. Consequently, it is recommended that both virus species-specific and genus or family-specific RT-PCR tests are used to ensure detection of virus species in the genus
Ilarvirus, Foveavirus, Trichovirus, Capillovirus and
Ampelovirus in
Prunus trees to overcome the risk of false-negative results, until improved tests can be developed. This study highlighted the need for further improvements in the development, selection and adoption of RT-PCR tests for detection of viruses in Australian grown
Prunus species.
HTS was used to confirm the first detection of CGRMV, CNRMV, LChV2 and PBNSPaV in Australia by RT-PCR and PCR. It was also useful to confirm the presence of viruses detected by a genus- or family-based RT-PCR test but not by the specific test (e.g., APCLSV). Metagenomics HTS detection of additional virus species, specifically ApLV, AVCaV, APV2, NSPaV and LChV1, and strains of CGRMV, CNRMV, LChV2 and PBNSPaV, which were not detected by species-specific or genus- or family-based RT-PCR tests, highlights the high potential for a false-negative diagnosis by RT-PCR due to virus genetic variability. It also highlights the value of metagenomics HTS as a non-discriminative diagnostic tool for virus detection compared to RT-PCR, because it does not rely on knowledge of the virus genome sequences [
77]. Given the high number of viruses that may need to be tested in some
Prunus species, which could make RT-PCR a cost-prohibitive diagnostic tool, the reducing cost of HTS [
28] makes this technology more attractive for routine virus detection in
Prunus species in Australia.
In this study, full genomes of Australian ACLSV, APCLSV, AVCaV, ApLV, APV2, NSPaV, PBNSPaV, CNRMV, LChV2, LChV1, CGRMV and CVA isolates were generated by HTS, contributing to the global knowledge of their presence and diversity. Except for ApLV and APV2, multiple genome sequences were obtained for the other viruses. Multiple full-genomes of CVA, CGRMV and LChV1 strains were also detected by HTS in individual
Prunus tree samples. Similar detection of multiple genetic strains or variants may also be achieved through traditional Sanger sequencing of cloned RT-PCR and PCR amplicons. However, limitations in the sequencing of a few clones due to time, labour and cost constraints may hamper the detections of some strains present in low titre mixed infections. This challenge in the detection of mixed virus genetic strains in a single plant has recently been overcome by HTS for other virus species [
48,
78,
79], which further highlights the sensitivity and applicability of HTS in characterising and detecting virus genetic strains.
Phylogenetic and sequence identity analysis of full virus genomes was used to determine the genetic relationship between Australian
Prunus virus isolates of ACLSV, APCLSV, AVCaV, ApLV, APV2, CGRMV, CNRMV, CVA, LChV1, LChV2, NSPaV and PBNSPaV from this study and isolates of each species previously published on GenBank. Where multiple strains of a species were identified, considerable genetic diversity was observed amongst the strains. As reported in other studies [
18,
72,
76,
80,
81,
82], these virus species were often distributed throughout their individual phylogenetic trees, rather than forming a single Australian cluster, and were related to overseas isolates originating from different regions. This suggests multiple introductions of each virus species into Australia. However, in some instances, for example, ACLSV, CVA, CGRMV, CNRMV, LChV1 and LChV2, clusters of some strains were observed, and this was most frequently associated with the cherry tree as a host. The original source of propagation material for each tree is not known, but it may be that where a cluster is observed in the same host, regardless of location, each of those trees was propagated from a single source tree and the genetic differences could be associated to evolution of the virus species with time. There is one exception, where two closely related LChV1 strains within a single cluster were from cherry and plum trees that were collected in Queensland and Victoria, respectively. This may be coincidental, or the two LChV1 strains may have a common origin. The common origin could be spread by a vector within an Australian
Prunus germplasm collection or an overseas germplasm, although no vector is known for LChV1. It could also be associated with the use of a common infected rootstock.
The multiple genetic strains of CVA, CGRMV or LChV1 within individual trees were highly divergent from each other and occurred in different phylogenetic clusters. For example, the three distinct CVA strains identified in sample TAS12 occurred in the three of the six major CVA clusters reported by [
83]. Co-infection of divergent virus strains is known [
84] and is likely to represent multiple infection events, which could be through grafting or transmission by an insect vector, although no vector is known for CVA, CGRMV and LChV1. The biological implication of infection with multiple distinct virus genetic strains is not known and further studies are required.
Recombination is a powerful evolutionary driving force that results in the generation of new genotypes in RNA viruses [
85,
86]. Recombination events have been previously reported in RNA viruses in the family
Betaflexiviridae and
Closteroviridae [
79,
84,
87,
88] and in this study, multiple recombination events were identified amongst Australian isolates of ACLSV, CGRMV, CVA, LChV1, LChV2 and PBNSPaV. All the Australian recombinant isolates shared similar phylogenetic groupings with their major parental sequence which indicates close genetic connectivity within the population of these recombinants. Previous studies have shown frequent recombination events between multiple divergent virus genetic strains occurring in individual plants [
79,
81]. Interestingly, no recombinants were observed in the six
Prunus trees that had multiple full-genome genetic strains of CVA, CGRMV and LChV1. This could indicate recent introductions of these viruses into Australia or recent multiple or single infection events.