Introduction

Ming aralia (Polyscias fruticosa L. Harms) is a perennial shrub (also considered a dwarf tree) native to tropical Asia that is commonly used as either an indoor or landscape ornamental. P. fruticosa belongs to the family Araliaceae, which includes several important foliage crops, including aralias and scheffleras [1]. In early studies, the presence of bacilliform viral particles was reported to occur widely and frequently among several members of this family, including scheffleras (Schefflera actinophylla and S. arboricola) and aralias (Polyscias balfouriana, P. balfouriana ‘marginata’, P. fruticosa and P. guilfoylei) [2]. In the same study, a mealybug-transmitted badnavirus, schefflera ringspot virus (SRV), was characterized and partially sequenced from schefflera plants. Bacilliform viral particles observed in the other members of the family Araliaceae were suggested to be isolates of SRV. However, PCR-amplified fragments of the isolates from schefflera and aralia did not cross-hybridize, indicating genetic heterogeneity among these isolates [2]. In 2017, virus-like symptoms were observed in P. fruticosa plants, viral particles were characterized, and a genomic sequence was obtained. This communication reports the complete sequence and genome organization of a new bacilliform virus from P. fruticosa and its relationship to other members of the genus Badnavirus.

Virus source and isolation

Ming aralia plants (P. fruticosa) showing virus-like symptoms were obtained from nurseries in the USA (Florida and California) and landscape plants in Nigeria. Leaf samples showed a mild mosaic differing from symptoms previously reported in this species (Fig. 1A). Virus particles were purified from symptomatic tissue as described previously for piper yellow mottle virus (PYMoV) and commelina yellow mottle virus (CoYMV) [3, 4]. Purified suspensions were examined by transmission electron microscopy (TEM) using samples mounted on carbon-coated formvar grids negatively stained with 2% phosphotungstic acid, pH 7.0 (PTA). Bacilliform virions characteristic of badnaviruses were observed by TEM in negatively stained purified leaf extracts from symptomatic but not asymptomatic plants (Fig. 1B). The particle size was, on average, 30 nm width and 120 in length, determined by measuring 100 individual particles (Fig. 1C).

Fig. 1
figure 1

A. Leaf showing common symptoms associated with the presence of PoMV. B. Leaf from an asymptomatic plant. C. Viral particles observed using TEM. D. Graphic representation of the genome of polyscias mosaic virus (PoMV). Orange boxes indicate the predicted open reading frames (ORFs), and conserved domains are indicated by overlaying blue boxes

Full size image

Genome sequencing

Preliminary sequence analysis using the degenerate primers Badna T+2 revealed low sequence similarity in the highly conserved RT+RNase region to previously reported badnaviruses, including SRV. However, a high level of similarity was detected among samples form the USA and Nigeria (98% identity within the RT-RNase H region). The complete genomic sequence was obtained using a sample from the USA. Briefly, genomic DNA was extracted from purified viral suspensions as described previously [3, 4]. Subsequently, the virus suspension was incubated with DNase I (4 units) and RNase A (25 g/mL) at 37 °C for 30 min, followed by digestion with proteinase K (200 µg/mL) in the presence of 1% SDS at 50 ° C for 30 min. The mixture was subjected to phenol:chloroform extraction, and the nucleic acid was precipitated using ethanol. DNA extracted from virions was used as a template for genome sequencing. Initially, the RT+RNase H region was amplified and sequenced with the degenerate primer pair BADNA-T (5’-MYM WNG CTC TGA TAC CA-3’) and BADNA-2 (5’-TAY ATH GAY GAY ATH YT-3’) [5]. The RT+RNase H sequence was used to generate a pair of outward-facing specific primers (AR2-L: 5’-CTC CTT GTG GAC TTG TCT TTG CAT AGA GG-3’ and AR2-R: 5’-GGA TGA ATG ATC AAG ACT GGA AGA TAG TGC-3’) designed to amplify the complete circular genome using Long Amp polymerase (NEB). PCR products were cloned using the pGEM-T Easy Vector System (Promega). The complete sequence was obtained from three clones by Sanger sequencing and primer walking. Sequences were assembled with Geneious R11 [6].

Sequence analysis and genome organization

The viral genome consists of a circular DNA molecule with a length of 7,592 bp and 39% G+C content. The sequence was deposited in the NCBI GenBank database under accession no. MH475918. The tRNA methylation site (5’-TGG TAT CAG AGC ATG GTT-3’), a highly conserved motif required for replication, was identified and set as the starting point of the sequence. Open reading frames (ORFs) were predicted using ORFfinder (NCBI) and Geneious R11 [6]. The genome contains three overlapping ORFs organized in a manner that is typical for badnaviruses (Fig. 1C), with overlapping stop/start codons between ORF1 and ORF2 (TTGA) and between ORF2 and ORF3 (ATGA). ORF1 (position 353 to 772 nt) encodes a putative protein of approximately 16.20 kDa. A domain of unknown function (DUF1319) restricted to members of the genus Badnavirus was identified, and the highest identity in ORF1 is 49% at the amino acid level with cacao mild mosaic virus (NC_033738). ORF2 (nt position 769 to 1,164) is predicted to encode a 14.19-kDa protein showing highest similarity (34% identity at the amino acid level) to its counterpart in cacao mild mosaic virus. ORF3 (positions 1,161 to 7,136) encodes a putative polyprotein of 228.46 kDa, with conserved domains identified including zinc finger, aspartic protease, reverse transcriptase (RT) and RNase H. The highest level of identity in ORF3 is 45% at the amino acid level with citrus yellow mosaic virus (EU708317). Pairwise alignments using only the RT+RNase H region showed the highest identity (63%) at the nucleotide level with schefflera ringspot virus (SRV). Nucleotide sequence divergence higher than 20% within the RT+RNAse H region is one of the criteria for species demarcation in the genus Badnavirus. Based on RT+RNase H sequence alignments between the P. fruticosa badnavirus and other members of the genus, this new virus should belong to a new species in the genus Badnavirus, and we propose the name “polyscias mosaic virus” (PoMV) for this virus. Analysis of the partial sequence obtained from the Ming aralia sample from Nigeria revealed a high level of identity to PoMV-USA (98%) at the nucleotide level, and therefore it was considered an isolate of PoMV (PoMV-NGA). The partial sequence of PoMV-NGA has been submitted to the GenBank database (MH475919).

Phylogenetic analysis

Phylogenetic relationships between PoMV and members of the genus Badnavirus were assessed using genetic information from 50 members of this genus. Trees were constructed using the complete genome sequences (Fig. 2A) and the RT+RNAse H region (Fig. 2B). Since the complete sequence of SRV is not available, it was not included in the complete genome analysis. Alignment of complete genome sequences was done using the MATFF algorithm in the Galaxy Platform [7], and maximum-likelihood (ML) phylogenetic analysis was carried out using the software MEGA 8.0 [8]. ML trees were inferred based on the general time-reversible model and gamma distribution with invariable sites [9]. The ML tree inferred from the complete genome sequences (Fig. 2A) locates PoMV in a single sister clade with cacao mild mosaic virus (KX276640), a recently reported badnavirus isolated from cacao. The ML tree inferred from the highly conserved RT+RNase H region (Fig. 2B) resulted in a tree with low overall bootstrap support; however, it clusters both isolates of PoMV in a clade with schefflera ringspot virus (SRV), cacao mild mosaic virus, sweet potato badnavirus B (NC_012728), and sweet potato pakakuy virus (NC_015655).

Fig. 2
figure 2

Maximum-likelihood phylogenetic trees of members of the genus Badnavirus. A. Tree based on the complete genome sequence. B. Tree based on the RT+RNase H region. Boxes and arrows indicate the position of PoMV. Trees are drawn to scale, with branch lengths measured in the number of substitutions per site. Bootstrap values (1000 replicates) are shown at each node

Full size image

SRV was first described and partially characterized from Schefflera actinophylla (sin. Brassaia actinophylla) [2]. In the same study, bacilliform viral particles were found infecting different members of the genus Polyscias (P. balfouriana, P. marginata and P. fruticosa). However, a genomic DNA-DNA hybridization assay revealed heterogeneity within the RT+RNase H region. The molecular characterization of PoMV supports previous observations about badnavirus diversity in the botanic family Araliaceae.