Alfalfa virus S, a new species in the family Alphaflexiviridae

A new species of the family Alphaflexiviridae provisionally named alfalfa virus S (AVS) was discovered in alfalfa samples originating from Sudan. A complete nucleotide sequence of the viral genome consisting of 8,349 nucleotides excluding the 3’ poly(A) tail was determined by high throughput sequencing (HTS) on an Illumina platform. NCBI BLAST searches revealed that the virus shares the greatest degree of sequence identity with members of the family Alphaflexiviridae, genus Allexivirus. The AVS genome contains six computationally-predicted open reading frames (ORF) encoding viral replication protein, triple gene block protein 1 (TGB1), TGB2, TGB3-like protein, unknown 38.4 kDa protein resembling serine-rich 40 kDa protein characteristic for allexiviruses, and coat protein (CP). AVS lacks a clear 3’ proximal ORF that encodes a nucleic acid-binding protein typical for allexiviruses. The identity of the virus was confirmed by RT-PCR with primers derived from the HTS-generated sequence, dot blot hybridization with DIG-labeled virus-specific RNA probes, and Western blot analysis with antibodies produced against a peptide derived from the CP sequence. Transmission electron microscopic observations of the infected tissues showed the presence of filamentous particles similar to allexiviruses in their length and appearance. To the best of our knowledge, this is the first report on the identification of a putative allexivirus in alfalfa (Medicago sativa). The genome sequence of AVS has been deposited in NCBI GenBank on 03/02/2016 as accession № KY696659.


Introduction
Alfalfa (Medicago sativa) is the most extensively cultivated forage legume in the world and the fourth most widely grown crop in the US, planted on more than 23 million acres in all 50 states. The estimated worth of alfalfa hay is over $8 billion annually. In addition to its value as livestock forage, alfalfa also has potential to make important contributions to bioenergy production. Alfalfa productivity is often limited by different diseases, insect pests and abiotic stress factors [1,2]. Minimizing these losses is a major area of concern in the alfalfa industry.
Although viral infections of alfalfa are widespread in major cultivation areas, they are generally considered to be diseases of minor or limited importance. The most frequently diagnosed viruses in alfalfa are Alfalfa mosaic virus and Cucumber mosaic virus [3]. Also, among the a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 different viruses infecting alfalfa are alfalfa enation virus, Bean leafroll virus, Lucerne transient streak virus, Pea streak virus, Red clover vein mosaic virus, Peanut stunt virus, Bean common mosaic virus, and others [4]. Alfalfa may serve as a natural reservoir for dissemination of viruses to other agriculturally important crops, although its exact role in the epidemiology of viruses in other crops is not well documented [5,6]. In recent years, emerging viral diseases of alfalfa with the capacity to cause more serious yield losses, have been described [7,8]. In addition, synergetic interactions of multiple viral infections can threaten alfalfa production [7,9].
In this work, we report identification, complete genomic sequence and specific diagnostics of a novel and unusual virus discovered in alfalfa samples originating from Sudan. The virus, provisionally named alfalfa virus S (AVS), was not previously described in alfalfa or any other plant species. Total RNA extraction, RT-PCR, dot blot hybridization and Illumina RNA sequencing For dot blot hybridization, RT-PCR assays and for Illumina RNA sequencing, total RNA was extracted from alfalfa samples using TRIzol RNA isolation reagent as described by the manufacturer (ThermoFisher Scientific). For dot blot hybridization, one μl of total RNA (~100-300 ng) was dotted directly onto a nylon membrane and cross-linked by ultraviolet light using a UVP HL-2000 HybriLinker (UVP, LLC, CA, USA) for three minutes. Hybridization (performed in UVP HL-2000 HybriLinker hybridization oven), washing and detection conditions were done exactly as described in the Boehringer Mannheim guide for filter hybridization [10].

Plant samples and inoculation of alfalfa plants
RT-PCR was performed with total RNA employing SuperScript RT-PCR system according to the manufacturer's directions (ThermoFisher Scientific). Primers for RT-PCR are shown in Table 1.
RT-PCR products were either sequenced directly or cloned into pCRII-TOPO vector with dual promoter (ThermoFisher Scientific) for sequencing and synthesis of DIG-labeled RNA probes. DIG-labeled probes were synthesized using DIG RNA labeling mixture and T7 RNA polymerase (ThermoFisher Scientific) essentially as described [10].
Library construction and high-throughput sequencing (HTS) on an Illumina Platform were outsourced and performed by SeqMatic Inc. (Fremont, CA). Contigs were assembled from the read data using CLC Genomics Workbench software version 9.5.2.
The 5'end of the virus was determined using the 5-RACE system (ThermoFisher Scientific), PCR with several genome-specific primers (LN416, LN421-424; Table 1) and sequencing. The 3' end of the genome was obtained by RT-PCR with primers LN412/LN427 and sequencing of the resulting PCR products.

Transmission electron microscopy
For transmission electron microscopy (TEM), leaves were homogenized in sterile water followed by centrifugation of the extracts in a bench-top Eppendorf centrifuge for five minutes at 16.1 k rcf (13.2 k rpm). Virus captured on the TEM grids was stained with 1% phosphotungstate (PTA) solution. The grids were examined in a Hitachi H-7700 Electron Microscope at the Electron and Confocal Microscope Unit, Beltsville Agricultural Research Center.

Nucleotide sequence and genome organization
Alfalfa plants collected in Sudan and used in this work exhibited chlorosis and stunting. Upon arrival, the samples were partially deteriorated; visual signs and symptoms for identification of any possible diseases were not clear enough to be recognized. Preliminary rounds of RT-PCR (not shown) with total RNAs extracted from the tissues suggested that plant sample designated 98.3A might be infected with several viral diseases that commonly occur in alfalfa. Total RNA from this sample was subjected to HTS to detect all potentially present viruses in an unbiased manner.
Raw data reads were assembled into contigs and searched against a custom virus database by the translated BLAST tool (BLASTX) [12]. Several essentially complete viral genomes were discovered in the sequenced sample: a novel flexivirus with a highest bit score for shallot virus X (ShVX), a virus with~98% identity to peanut stunt virus (PSV, genus Cucumovirus, family Bromoviridae), and a virus with 90-97% identity to alfalfa enamovirus-1 (AEV-1, tentative member of the family Luteoviridae), [7]. To confirm the identity of the sequenced sample as Medicago sativa species, all host transcripts were assembled into a separate dataset and mapped onto the reference genome of cultivated alfalfa (inbred lines of cultivated alfalfa at the diploid  [7,13], AVS is a new, previously undescribed species. A complete nucleotide sequence of the viral genome obtained by de-novo assembly of HTS-generated reads, 5'RACE and sequencing of the RT-PCR-amplified 3'terminus, consisted of 8,349 nucleotides excluding the 3' poly(A) tail. NCBI BLAST searches revealed that the virus has its greatest sequence identity with members of the family Alphaflexiviridae, genus Allexivirus. On the nucleotide level, the species most similar to the AVS were isolates A and B (54.6% and 53.7% identities, respectively) of unclassified arachis pintoi virus [14]; an unassigned species within the family Alphaflexiviridae, blackberry virus E (BVE) [15], (50.6%); and the type member of the genus Allexivirus, ShVX [16], (49.7%). Nucleotide identities were obtained by sequence-based virus classification tool, Pairwise Sequence Comparison (PASC) [17].

Phylogenetic analysis
A phylogenetic tree was generated with the complete nucleotide sequences of all currently known allexiviruses, closely related unclassified members of the family Alphaflexiviridae and the AVS genome (Fig 2). AVS grouped with arachis pintoi virus A (54.6% identity vs AVS), branching out of the larger BVE group (50.6% identity to AVS), which is directly connected to the cluster of established Allexivirus species.

Confirmation of the identity and specific detection of the virus
A complete genome of the AVS obtained via RNA-sequencing was used to design several sets of primers for RT-PCR to confirm the virus identity and accuracy of the nucleotide sequence. Using primer pairs for CP ORF and p38.4 ORF (Table 1) with total RNA extracted from the original infected sample 98.3A, respective virus genes were successfully amplified (Fig 3), cloned into pCRII-TOPO vector (ThermoFisher Scientific) and sequenced. Sequence data completely matched the HTS results thus validating the quality of the digital assembly and the presence of the predicted virus.
To further support the existence of the pathogen in the defined molecular form, DIGlabeled RNA probes were generated from the cloned p38.4 and CP cDNAs and used in dot blot hybridization assays with RNA extracted from infected alfalfa tissues. Both RNA probes successfully detected virus in the infected samples 98.3A, 98.1A and 98.3A(0), which are different alfalfa plants belonging to the same cultivar (Fig 4).
Molecular data were also supported by immunological detection of the virus: polyclonal antibodies, produced against the peptide CP1 within the AVS CP specifically reacted with alfalfa samples 98.3A, 98.1A and less obviously with the sample 98.3A(0) in WB assay (Fig 5). The CP migration in the 10-20% SDS gel slightly differed from the predicted protein size (~29 kDa vs 31kDA, respectively). The faster migration can probably be explained by atypical protein-SDS binding, albeit leaky scanning and initiation of the CP translation from alternative downstream codons remains a possibility. The same three samples were tested positive in dot blot hybridization assays with the RNA probes to AVS CP and p38.4 (Fig 4). Antibodies to peptide CP2 reacted less specifically, while Abs to peptide CP3 did not react at all (not shown). Thus, western blot analysis suggested the presence of antigenicity near the N-terminal region of the AVS CP.

Transmission electron microscopy
The original sequenced sample 98.3A used for HTS analysis and also two more samples, 98. 1A and 98.3A (0), that were later confirmed to be positive for AVS by dot blot hybridization and WB, were examined in a Hitachi H-7700 Electron Microscope. Filamentous virus particles about 600-800 nm in length and 13-16 nm in diameter were observed in the extracts from all tested samples (Fig 6). Shorter and longer filaments possibly representing broken particles and end-to-end aggregation of two virions, respectively, were also observed (not shown).

Discussion
As of today, the natural host range of all ICTV-classified allexiviruses is thought to be restricted to Allium spp. of monocotyledonous flowering plants, although some allexiviruses will induce local lesion on Chenopoidium species [16,18,19]. Here we report on the discovery, characterization and sensitive detection of a novel viral pathogen, resembling allexiviruses, in  dicotyledonous species Medicago sativa, legume family Fabaceae. The virus, tentatively named "alfalfa virus S" (AVS), is an unusual pathogen most closely related to arachis pintoi virus and BVE, which are putative and unclassified members of the family Alphaflexiviridae, respectively, and to several members of the genus Allexivirus. Even though the identity level between AVS and accepted allexiviruses is rather low, it has some of the characteristic features only found in allexiviruses, such as a serine-rich p38.4 protein of unknown function, as well as its general genome organization. Phylogenetic analyses clearly placed AVS together with arachis pintoi virus and BVE in a distinct cluster associated with the identified allexiviruses (Fig 2). However, unlike arachis pintoi virus (lacks TGB3) and BVE (lacks an AUG initiation codon of TGB3), the putative TGB3-like protein of the AVS has an AUG initiation codon, suggesting the functional importance of TGB3 rather than any alternative strategy for the viral cell-to-cell movement [15]. Similarly to arachis pintoi virus and BVE, no sequences homologous to the 3' proximal nucleic acid-binding protein of allexiviruses were identified in the assembled genome of AVS, perhaps suggesting the need for a new genus, accommodating these viruses.
The identity of the virus was confirmed by RT-PCR with primers derived from the HTSgenerated sequence, dot blot hybridization with DIG-labeled virus-specific RNA probe, and WB analysis with antibodies produced against a viral CP peptide. Each of these methods can be of practical use for accurate diagnosis, identification and control of AVS in single infection or in complex multiviral epidemics of alfalfa. TEM observations have shown a presence in the infected tissues of filamentous virions similar to allexiviruses in their length and appearance.
We conclude that AVS represents a unique virus species with a general resemblance to the members of the genus Allexivirus and a more evident relation to the unclassified arachis pintoi virus and unassigned alphaflexivirus BVE. However, a low level of the whole genome homology with the most closely related viral species (49%-54%) indicates that AVS is a member of a novel, previously undescribed genus.