Horse nettle (Solanum carolinense L.) is a herbaceous perennial commonly found on roadsides and in agricultural fields. The plant propagates by seeds and overwintering creeping rhizomes, which give it a growth advantage over other weed species. This together with the fact that very few herbicides are effective at controlling it, raises the concern that this noxious weed could serve as a virus reservoir [1, 2]. In 2021, a horse nettle showing yellowing symptoms was collected in an apple orchard located at Beltsville, Maryland, U.S. Total RNA was extracted from leaves using the RNeasy Plant Mini kit (Qiagen, Valencia, CA); cDNA libraries were prepared with the TruSeq Stranded Total RNA Ribo-Zero Plant kit and sequenced on the NextSeq 500 platform (Illumina, San Diego, CA) with the 75-cycle high output sequence kit (1x 86bp, single-end), generating 30,554,603 reads. Virus detection was performed using the USDA-APHIS PGQP PhytoPipe pipeline [3]. Briefly, raw reads were filtered and trimmed using Trimmomatic (v.0.39) [4]. Then, contigs were assembled using Trinity (v2.8.6) [5] and were compared to the NCBI viral reference database (Jan 2022) by BLASTn search [6], and to the Reference Viral Database (RVDB) protein database (v22) [7] by Diamond blastx search [8]. Two contigs (7284bp and 4286bp) were found to be 55.68% identical to tomato black ring virus (TBRV) RNA1 (UBZ25932) and 44.77% identical to potato virus B (PVB) RNA2 (QXO36679) at the amino acid level, respectively. After mapping reads to the two contigs using BWA (v0.7.17-r1188) [9], a high level of coverage (6340 and 4345X) was attained. The two contigs were confirmed by RT-PCR and Sanger sequencing of 18 regions (see primers in Supplementary Table 1). The 5’ and 3’ ends of the virus genome were amplified using SMARTer RACE 5’/3’Kit (Takara Bio USA, San Jose CA) and virus-specific primers (Supplementary Table 2) according to the manual. The complete genome sequence of the virus, tentatively named as horse nettle virus A (HNVA), is comprised of two RNA molecules: RNA1 (7522 nt; Genbank accession OP292294) and RNA2 (4710 nt; Genbank accession OP292295) excluding poly(A) tails (Fig. 1).
RNA 1 contains a single ORF (6897 nt) coding for a polyprotein (257.92 KDa), which shares maximum identity (55%) with tomato black ring virus (TBRV) polyprotein. Analysis of the molecule revealed conserved motifs of multiple coding products of nepovirus including: 1) protease cofactor (F-x27-W-x11-L-x21-L-x-Q, x represents random amino acid) between 439-502 nt; 2) motif A (G-x4-GKS, 800-807 nt), motif B (D, 850 nt), and motif C (N, 897 nt) of NTP-binding helicase followed by 3) consensus sequence of VPg (RN-x5-R) between 1244-1251 nt; 4) a Serine protease characteristic motif (H-x37-E-x94-C/SG-x8-G-x5-G-x-L-x2-G) between 1296-1450 nt, and 5) RdRp core domain (D-x4-D-x59-G-x3-T-x3-N-x33-GDD) located at 1750-1859 nt. Pairwise alignment conducted between HNVA and subgroup B nepoviruses identified five dipeptides (RA394-395, KS625-626, KS1231-1232, RS1256-1257 and KA1469-1470) as the cleavage sites of the polyprotein. Five putative coding products include X1, X2, NTB, VPg, protease, and RdRp with size of 45.1KDa, 25.2KDa, 66.8KDa, 3.1KDa, 24.1KDa and 93.8KDa, respectively.
RNA2 also consists of a single ORF (222-4283 nt) coding for a polyprotein (150.3 KDa) sharing maximum identity with potato virus B (PVB) (45%). Conserved motifs of nepovirus movement protein (MP, LPL) and coat protein (CP, F-x-FYGRS) were identified at 637-639 aa and 1335-1341 aa, respectively. Putative CP (56.9 KDa) is separated from MP at C838/S839, however, a cleavage site between MP and the putative protein 2A was not identified. Interesting enough, part of the RNA2 molecule (2850-4259 nt) is predicted to be a genetic recombinant derived from beet ringspot virus (BRSV, major parent) and PVB (minor parent), both of which belong to the subgroup B nepovirus. Recombination events were identified using RDP v.5.5 [10] by at least three detection methods with an associated p value of <0.05 and with at least one method having an associated p value of <0.001 coupled with supporting phylogenetic evidence, as mentioned previously [11, 12] (Supplementary Table 3).
The 5’-UTRs of RNA1 and RNA2 are 198 nt and 221 nt in length, respectively. Although their sequence identity is only 61%, they do share the same 5’-terminal nucleotides (UUGAAAAUUUCUGUGAA), which is identical to the consensus sequence of the 5’ end of nepovirus genome [13]. The length of 3’-UTRs of the two RNA molecules is the same (427 nt) and their sequences are highly identical to each other (98%), as reported for other nepoviruses in the subgroup B [14, 15]. Phylogenetic analysis of the amino acid sequences of nepovirus CPs and protease-polymerase (Pro-Pol) regions indicated HNVA form the group with those of the subgroup B of the genus (Fig. 2). Given that HNVA shares minimal level of identity with any known members of the genus Nepovirus for both CP (52%) and Pro-Pol (63%), which is far below the species demarcation criteria, it should be classified as a novel member of subgroup B nepovirus. Preliminary nematode screening carrying this novel virus along with taxonomic classification suggested the potential vector to be Xiphinema rivesi, which were isolated from soil and roots of the symptomatic horse nettle plant. Further research is needed to ascertain the transmissibility of the virus as well as to study the virus population diversity.