Snake River alfalfa virus, a persistent virus infecting alfalfa (Medicago sativa L.) in Washington State, USA

Abstract

Here we report an occurrence of Snake River alfalfa virus (SRAV) in Washington state, USA. SRAV was recently identified in alfalfa (Medicago sativa L.) plants and western flower thrips in south-central Idaho and proposed to be a first flavi-like virus identified in a plant host. We argue that the SRAV, based on its prevalence in alfalfa plants, readily detectable dsRNA, genome structure, presence in alfalfa seeds, and seed-mediated transmission is a persistent new virus distantly resembling members of the family Endornaviridae.

Main text

Snake River alfalfa virus (SRAV) was recently identified from alfalfa plants and thrips Frankliniella occidentalis collected in the Minidoka and Twin Falls counties of Idaho, USA [1]. Based on the genome structure and phylogeny of its RNA-dependent RNA polymerase (RdRp), the authors hypothesized that SRAV is the first flavi-like virus identified in a plant host [1]. The SRAV polyprotein, however, contained no predicted helicase domain found in all flaviviruses.

To date, no occurrences of SRAV have been reported in alfalfa or on different hosts from other locations. In this work, applying high-throughput sequencing (HTS), we detected SRAV in 50 individual alfalfa plant samples collected from 10 commercial fields in Grant County, WA. Plants used for RNA extraction displayed a diverse symptomatology that occasionally correlated with the symptoms allegedly reported for SRAV, such as yellowing and vein clearing (Fig. 1), [1]. These symptoms, however, were likely due to the presence of multiple co-infecting pathogens in the same plants.

Fig. 1

Examples of different symptomatology observed on the leaves of alfalfa (Medicago sativa L) plants infected with Snake River alfalfa virus (SRAV-WA1) in Washington State, USA. Sample 1 contained 7,016 reads of SRAV and 11,797 covered bases; sample 2 contained 10,684 reads mapped to the virus and 11,699 covered bases; sample 3 contained 47,175 reads mapped to the virus and 11,838 covered bases; sample 4 contained 71,267 viral reads and 11,811 covered bases; sample 5 contained 8,079 reads and 11,872 covered bases (100% vs. reference genome); and sample 6 contained 19,539 viral reads and covered 11,795 bases. Multiple viral, fungal and bacterial pathogens were also identified in samples in which SRAV was found. These included alfalfa mosaic virus, bean leaf roll virus, pea streak virus, partitiviruses, amalgavirus, lucerne transient streak virus, Alternaria alternata, Alternaria arborescens, Bipolaris spp., Stemphylium lycopersici, Fusarium spp., Pseudomonas spp., Erwinia spp., etc. thus emphasizing the importance of pathobiome in signs and symptoms of disease

Total RNA was extracted using Promega Maxwell® RSC Plant RNA Kit (Promega Corp., Fitchburg, WI). Library preparation was performed with Illumina TruSeq Stranded Total RNA with Ribo-Zero kit (Illumina Inc., San Diego, CA), and the sequencing platform used was HiseqX10 (PE150) (Omega Biosciences, Norcross, GA). Bioinformatic pipeline included adapter trimming followed by de-novo assembly of reads, unmapped to M. sativa genome and M. truncatula mitochondrion genomes using SPAdes [7] in HMM-guided mode. Phylogenetic analysis was performed with MEGA software [6] using Maximum Likelihood method and bootstrap analysis of 1000 replicates. Conserved RdRp domains for multiple alignment were obtained using InterPro tool (https://www.ebi.ac.uk/interpro/).

All 50 unique alfalfa plant samples contained viral reads, varying in quantity from 46 to 71,267 (Table 1). Total number of reads mapping to the SRAV genome was 1,017,715 with an approximate length of each read 150 nt (Table 1). Several contigs apparently representing the complete genome of the virus were assembled de novo. The sequences varied slightly in length and contained small number of SNPs, indicating a presence of different genetic variants of the virus. Nevertheless, the sequence identity when compared to the reference genome (ON669064) in all cases remained > 99%.

Table 1 Occurrence of SRAV-related reads in 10 commercial alfalfa fields (five samples per field) of Grant County, Washington State

One of the de novo-assembled contigs recovered from the individual library and sample containing 8,079 viral reads, was 11,811 nt in length and had 100% base coverage with the reference, thus representing a complete genome of the virus. It was 7 nt longer at the 5' end than ON669064, which was confirmed by 5' RACE using SMARTer® RACE 5'/3' Kit (Takara Bio USA, Inc., San Jose, CA). The 3’ end of the sequence was 59 nt longer than that of the reference ON669064 and matched another isolate of the virus, SRAV_ALF1071, found in GenBank under accession number ON669090.1 [1]. Application of the 3'RACE also showed that the virus has ~ 30-long 3′-terminal poly(A) tract, which is absent in all members of the family Flaviviridae ([14]; https://ictv.global/report/chapter/flaviviridae/flaviviridae).

At the nucleotide level, the SRAV-WA1 (for Washington State) was 99.8% identical to the reference genome ON669064 with 18 SNPs between the two, therefore depicting an isolate of the same virus. The genome of SRAV-WA1 encoded a single 3,835 amino acid (aa) polyprotein 99.9% identical to the reference. BLASTP and PSI-BLAST searchers of the SRAV-WA1 polyprotein against the GenBank database identified no related sequences except existing SRAV submissions. In silico analyses of the viral polyprotein using Pfam, InterPro and CDD databases revealed the presence of the conserved RdRp domain (3220–3478 aa, E-value = 1.43E-21, InterPro). No other domains were reliably identified. A weak relation to the superfamily of trypsin-like serine proteases (E-value = 1.01e-03) was found in the 1855–1910 aa region of the polyprotein when Superfamily database (https://supfam.org) was used to detect protein sequence similarities.

The results obtained by HTS were validated by RT-PCR with two sets of primers in three technical replicates using SuperScript III One-Step RT-PCR System (Thermo Fisher Scientific, Waltham, MA). One set was ANPV_3 derived from Dahan et al. [1], and another set of primers was designed in this work: LN1036-F, GGGAGAACCAGGAAACTGTTAG and LN1037-R, CTGTCGCATAGTCCGCTTATT. RT-PCR using both primers pairs produced correct amplicons, while no products were generated from control samples in which no SRAV-related HTS reads were found (Fig. 2). The amplicons were sequenced and validated to be SRAV.

Fig. 2

RT-PCR with primers specific for Snake River alfalfa virus. M, 1 kb plus DNA marker (Thermo Fisher Scientific, Waltham, MA). Lanes 1,2: RT-PCR products amplified with primers LN1036/37 and ANPV_3, respectively. Lanes 3,4 and 5: amplification from alfalfa samples containing no SRAV reads, water control, and Taq DNA polymerase (no RT mix added to verify the absence of genomic DNA), respectively; primers LN1036/37. Lanes 6, 7, and 8: representative RT-PCR products amplified from seeds of alfalfa cultivars SW-9215, CUF101, and Maverick using LN1036/37 primers. Lanes 9,10, and 11: the same reaction controls as shown in lanes 3–5

The omnipresence of the virus in all analyzed samples indicated the possibility of persistent infection like those caused by partitiviruses and endornaviruses [16]. Considering resemblance in the size and structure of the genome, we compared SRAV to endornaviruses. Viruses in this family infect plants, fungi, and Oomycetes, and are generally associated with symptomless infections and no pathogenic effects [3]. They have a linear genome of 10 to 17 kbp in length, that encodes a polyprotein ranging from 3,217 to 5,825 aa [3, 16, 19]. Notably, several known endornaviruses, same as SRAV, lack helicase domain [16, 19]. While members of the Endornaviridae family were often reported as double-stranded RNA viruses [3, 16], current ICTV classification describes them as single-stranded, positive-sense RNA genomes that have been characterized using replicative dsRNAs forms [19].

Phylogenetic analysis using the polyproteins of SRAV, different viruses of the family Endornaviridae, and members of the family Flaviviridae, placed both SRAV isolates within Endornaviridae, although SRAV isolates formed a separate cluster (Additional File 1). When we performed phylogenetic analysis using InterPro-extracted RdRp domains of the Endornaviridae and Flaviviridae (3,204–3,462 aa in SRAV-WA1), SRAV clustered with the former as well, again forming a distinct grouping (Additional File 2). It is worth noting, however, that SRAV placement was not consistent, pointing to potentially incorrect phylogenies or an irreproducibility in maximum likelihood inference and [18].

When we used Sequence Demarcation Tool program that allows classification of virus sequences based on sequence pairwise identity (SDTv1.2; [10]), it showed low similarities of both polyprotein and RdRp of SRAV with those of endornaviruses and flaviviruses (Additional Files 3 and 4).

Since plant endornaviruses are transmitted through seeds via the gametes [8, 13, 19], we decided to test seeds of several alfalfa cultivars for the presence of the virus by RT-PCR. Seeds were scarified with concentrated H₂SO₄, surface-sterilized with 70% ethanol, and rinsed with sterile water [11]. Total RNA was extracted with Takara Plant and Fungal RNA isolation kit (Takara Bio, San Jose, CA) and used in RT-PCR with primers LN1036/37. RT-PCRs with five out of six tested seed samples derived from different alfalfa cultivars (Maverick, SW-9215, SW-8421, SW-9720, and CUF101) were virus-positive, indicating a high rate of seed infection (Fig. 2). Resultant amplicons were sequenced and confirmed to be SRAV-WA1. Seeds of one cultivar, Regency SY, were RT-PCR-negative (not shown). To additionally confirm seed transmission of the virus, leaves of the germinated seedlings were randomly checked by RT-PCR one week after germination. Except for Regency SY, seedlings of other tested cultivars were positive for SRAV-WA1 (not shown). These experiments demonstrated localization of SRAV-WA1 in the internal parts of the seed, likely in the embryo. They also showed a high rate of seed infection by the virus, and its efficient vertical transmission via seeds, thus confirming persistent nature of the virus [15].

One of the characteristic features of all endornaviruses is readily detectable viral replicative form, double-stranded RNAs (dsRNAs), that accumulates in the host tissues in high quantities [19]. To extract dsRNA from leaves of the SRAV-WA1-infected alfalfa plants, we followed the protocol of Khankhum et al. [5]. Agarose gel electrophoresis showed the presence of dsRNA of the approximately correct size corresponding to that predicted by de novo assembly of the HTS reads (Fig. 3).

Fig. 3

Agarose gel electrophoresis of dsRNAs extracted from leaves of two different alfalfa plants containing SRAV-WA1 reads. M, 1 kb plus DNA ladder (Thermo Fisher Scientific, Waltham, MA). Lane 1, dsRNA extracted from the sample A2-2 (Table 1). Lane 2, dsRNA extracted from the sample B8-2 (Table 1). Arrow indicates the predicted dsRNA of SRAV-WA1

Additionally, given that many endornavirus RNAs have a site-specific discontinuity (nick) on 5’ terminus of the coding strand, we have attempted, but failed, to determine its presence in the genome of SRAV-WA1 using 5'RACE approach [13].

Dahan et al. [1] detected SRAV in western flower thrips and suggested a possible role for the insect in virus transmission. However, thrips are known to transmit tospoviruses and plant viruses in the Ilarvirus, Carmovirus, Sobemovirus and Machlomovirus genera [4]. Based on our data, SRAV, analogously to vertically transmitted endornaviruses [8, 9, 13, 19], is also transmitted by seeds. Although alfalfa is one of the primary hosts for western flower thrips (and other species) and acquiring the SRAV during feeding cannot be excluded, transmission of the virus by thrips would require additional experimental confirmation.

Other viruses found in samples infected with SRAV-WA1 included alfalfa mosaic virus, pea streak virus, lucerne transient streak virus, bean leaf roll virus, partitiviruses, and amalgavirus. Fungal and bacterial pathogens, described in alfalfa, like Alternaria spp., Bipolaris spp., Stemphylium spp., Fusarium spp., Pseudomonas spp., Erwinia spp. etc. were also detected. These findings suggested that traditional Koch’s postulate of “one microbe—one disease” should be broadened into the principle of a pathobiome, when disease symptoms are attributed to a diverse community of pathogenic organisms within the plant, rather than to a single infectious agent [20].

Overall, our research confirmed association of SRAV with alfalfa and, for the first time, identified an extensive occurrence of this virus in Washington State. The importance of this work also relies on the hypothesis that placement of SRAV within the flavi-like lineage, as suggested by Dahan et al. [1], may not be entirely accurate. Prevalence of the virus in alfalfa plants, its genome organization, seed-mediated transmission, presence of the easily detectable dsRNA and, although partly, phylogenetic reconstruction, suggest that SRAV is a persistent virus possessing some features characteristic for endornaviruses.

However, the low percent identity of SRAV with endornaviruses and flaviviruses, absence of the poly (C) and presence of the poly (A) tract at the 3’ terminus of the genome, and lack of the site-specific nick at the 5’ end, indicate that SRAV may represent an entirely new taxonomic group of persistent viruses that does not belong to either of the two families. Altogether, more data are needed to assess taxonomy, biology, and economic importance of the virus.