In October 2017, several Fusarium strains were isolated from ginseng (Panax ginseng) root rot tissues from various locations in Jilin Province, China. These strains included F. solani SJH 2-4 from Songjianghe County, Baishan City; F. solani TH 7-2 from Tonghua City; F. oxysporum FS 7-7-3 from Fusong County, Baishan City; F. verticillioides BS 1-7 from Baishan City, and F. proliferatum DH 8-9 from Dunhua City. These strains were cultured on potato dextrose agar (PDA) medium (20 g/L glucose, 200 g/L potato, and 15 g/L agar) at 28°C. DNA was extracted using standard DNA extraction kits, and the identity of each strain was confirmed through polymerase chain reaction (PCR) amplification with universal (ITS1 and ITS4) and specific (EF-1 and EF-2) primers (O’Donnell et al., 1998; Bellemain et al., 2010). Phylogenetic trees were constructed, and the classification of each strain was further supported by morphological analysis. Details of all primers used in this study are provided in Supplementary Table S1.

For total RNA extraction, each strain was cultured on cellophane membranes placed over PDA for 4–5 days. Approximately 0.1 g of fresh mycelia was harvested, pulverized in liquid nitrogen, and processed using a total RNA extraction kit (TIANGEN, Beijing, China).

Double-stranded RNA (dsRNA) was extracted according to the method described by Zhao et al. (2020), with minor modifications. One g of young mycelium that had been cultivated on PDA for 4 days was flash-frozen in liquid nitrogen and ground into a powder. This powder was then transferred into a 10 mL sterile centrifuge tube, to which 2 mL of GPS buffer (glycine 15.0 g/L, Na₂HPO₄ 14.2 g/L, NaCl 35.1 g/L, pH 9.6), 2 mL of phenol (pH 8.0), 2 mL of chloroform isoamyl alcohol (24:1), 100 μL of 10% SDS, and 60 μL of β-mercaptoethanol were added. The mixture was shaken, homogenized, and incubated on ice for 30 min. Following centrifugation at 12,000 rpm for 10 min at 4°C, the supernatant was transferred to a new tube. Anhydrous ethanol was added to achieve a final ethanol concentration of 16%, followed by 0.1 g of cellulose powder CF-11 (Sigma Aldrich, St. Louis, MO, United States); it was then mixed and left to stand in an ice bath for 30 min. The dsRNA was precipitated using isopropyl alcohol and sodium acetate, followed by washing with 75% ethanol, and was ultimately dissolved in 30 μL of RNase-free water. This dsRNA was then treated with enzymes including S1 Nuclease and Recombinant DNase I (TaKaRa, Dalian, China) before electrophoresis on a 1.0% agarose gel.

For RT-PCR detection, first-strand cDNA was synthesized using GoScript Reverse Transcriptase (Promega, United States), followed by PCR amplification with specific primers (2-4F and 2-4R), as detailed in Supplementary Table S1.

A cDNA library was constructed using the TruSeq™ RNA Sample Preparation Kit, following the manufacturer’s guidelines. Paired-end sequencing of the barcoded library was then performed on the Illumina HiSeq X Ten platform at Shanghai Biotechnology Co., Ltd. Rapid amplification of cDNA ends (RACE) was used to obtain the 5′ and 3′ ends of the viral RNA sequence. The desired PCR products were purified, cloned into the pMD18-T vector (TaKaRa, Dalian, China), and verified by PCR using specific primers (M13-47, M13-48, and RACE primers). Plasmid extraction was then performed, followed by sequencing at Sangon Biotech. The primers used for cDNA cloning and sequencing are provided in Supplementary Table S1. Secondary structure predictions of the 3′- and 5′-termini of the viral RNA were conducted using the UNAfold web server².

The complete cDNA sequences of viruses present in the F. solani strain SJH 2-4 were retrieved from the NCBI GenBank databases and compared with previously reported viral sequences. Potential open reading frames (ORFs) and associated proteins were predicted using DNAMAN 7.0 software and analyzed using the Motif Search Tool³. Amino acid sequence alignments were performed using Clustal X software to ensure accuracy and consistency. Phylogenetic relationships based on viral RdRp sequences were analyzed using MEGA 7.0 software and the maximum likelihood method.

Protoplast regeneration was employed to remove FsoOLV1 following the method of Wang et al. (2020) with minor modifications. Protoplasts from strain SJH 2-4 were prepared and plated on RM medium, followed by incubation at 28°C. Mycelial tips that emerged were subsequently transferred to PDA medium supplemented with 0.3 mg/mL ribavirin for 10 generations of continuous culture. Total RNA was extracted from the transformants, and the elimination of the virus was verified using RT-PCR with virus-specific primers.

To assess the vertical transmission of FsoOLV1, 21 individual spores were isolated from the parent strain SJH 2-4 using single-conidium isolation. The conidia were cultured on PDA for 5 d, and RT-PCR was performed to detect FsoOLV1, using primers 2-4F/R (Supplementary Table S1), which amplified a 763-bp product. For horizontal transmission, co-cultivation experiments were conducted. SJH 2-4 served as the donor strain, and TH 7-2 and 2-4 VF1 were used as recipient strains. The strains were co-cultured on PDA plates until the colonies intersected, and mycelial blocks from TH 7-2 and 2-4 VF1 were transferred for further culturing and purification. The presence of FsoOLV1 was confirmed by RT-PCR in the purified strains.

PEG-mediated transfection was used for viral RNA transfection (Yu et al., 2010). Protoplasts from F. oxysporum strain FS 7-7-3, F. verticillioides strain BS 1-7, F. proliferatum strain DH 8-9, and F. solani strain TH 7-2 were prepared and mixed with total RNA from F. solani strain SJH 2-4. The mixture was incubated with PEG buffer (PEG4000 600 g/L, Tris-HCl 50 mM/L, CaCl₂ 50 mM/L, pH 7.5), spread on RM medium, and incubated at 28°C to facilitate growth. Once transformants emerged, they were subcultured onto PDA medium and subjected to RT-PCR for viral transmission assessment.

The colony morphology and mycelial diameter of the strains were estimated on PDA plates after 3, 5, and 7 days of inoculation using the 9 mm mycelial plugs. To assess the effect of the virus on mycelial biomass, the mycelia cultured on PDA were scraped off, dried at 80°C for 1 h, and weighed to determine mycelial biomass. The influence of the virus on hyphal morphology was examined by placing hyphal plugs on Spezieller Nährstoffarmer Agar (SNA) medium (KH₂PO₄ 1 g/L, KNO₃ 1 g/L, MgSO₄⋅7H₂O 0.5 g/L, glucose 0.2 g/L, sucrose 0.2 g/L, agar powder 20 g/L) in Petri dishes, followed by incubation at 25°C for 10 d. The morphology of the hyphal tips was observed under a microscope. To investigate the effect of the virus on conidia production and morphology, mycelial plugs (9 mm in diameter) were taken from the periphery of 3-day-old colonies on PDA plates, with two plugs per strain, and inoculated into a medium containing 100 mL of carboxymethylcellulose sodium (CMC) (carboxymethylcellulose sodium 15 g/L, NH₄NO₃ 1 g/L, NH₂PO₄ 1 g/L, MgSO₄∙7H₂O 0.5 g/L, and yeast 1 g/L) within a 500 mL flask. This procedure was replicated three times for each strain. The flasks were then placed in a shaker incubator at 25°C (180 rpm). After 3, 5, and 7 days of culture, the liquid medium was filtered using three layers of sterile wipes. The conidia count for each strain was conducted using a hemocytometer, and spore morphology was examined under a microscope; date from 7 days were used for analysis. To evaluate the effect of the virus on spore germination rates, the spore suspensions from each strain cultured in CMC were diluted to a uniform concentration (1 × 10⁷), and 1 mL of this conidial suspension was transferred to 100 mL of YEPD (yeast 3 g/L, peptone 10 g/L, glucose 20 g/L). After incubation at 200 rpm and 25°C for 10 h, the germination rate was assessed under a microscope, and spore morphology was observed. The experiments were independently replicated three times. Data analyses were conducted using SPSS software (version 20.0, IBM Corp.). A one-way analysis of variance (ANOVA) was employed to compare the datasets, followed by Tukey’s honestly significant difference (HSD) test for post hoc analysis. The threshold for statistical significance was P < 0.05.

The virulence of the fungal strains was evaluated through direct inoculation on Panax ginseng roots. Healthy, 3-year-old Panax ginseng “Damaya” roots were used, and three 1 mm deep holes were created using an inoculation needle. Fungal plugs were then placed onto the wounds, and observations and photographs were taken after 7 days. Control Panax ginseng roots were inoculated with pure PDA medium. Lesion sizes were measured at 7 days post-inoculation. Three roots from each group were selected for the experiment, and three replication of this experiment were performed. Statistical analysis was performed using the same method as described previously.

Total RNAs were extracted from six F. oxysporum strains, one F. solani strain, and one F. proliferatum strain, all isolated from ginseng (Panax ginseng) root rot samples, and subsequently pooled for next-generation sequencing. Nine viral contigs were identified. Using BLASTX on NCBI, one of these viral contigs, contig 30, showed similarity to Plasmopara viticola lesion-associated ourmia-like virus 13 (QGY72543). RT-PCR analysis confirmed the presence of contig 30 in the F. solani strain SJH 2-4, which was isolated from Songjianghe city in Jilin Province, China. Further RT-PCR detection of nine viral contigs revealed that SJH 2-4 contains only contig 30. Based on these findings, we named this mycovirus as “Fusarium solani ourmia-like virus 1” (FsoOLV1).

Although the colony morphology of SJH 2-4 on PDA medium was similar to that of the virus-free F. solani strain TH 7-2, its growth rate was significantly slower (Figures 1A, B). Virulence assessment revealed that SJH 2-4 caused small lesions on Panax ginseng roots, in contrast to the severe root rot caused by the virus-free TH 7-2 strain (Figures 1C, D). Additionally, the fragment size of dsRNA extracted from strain SJH 2-4 and treated with DNase I and S1 nuclease was approximately 3.0 kb (Figure 1E).

The full-length cDNA sequences of FsoOLV1 were determined by cloning and assembling cDNAs obtained from RT-PCR and RACE (Supplementary Figure S1). The complete sequence of FsoOLV1 spans 2,801 nucleotides (nt) and was submitted to GenBank under accession number OP807121. The genome of FsoOLV1 has a GC content of 47.05% and contains a single ORF that encodes an RdRp consisting of 750 amino acids, with a molecular weight of approximately 84.84 kDa. This ORF starts at an AUG start codon at position 55 and ends at a UAG stop codon at position 2,310 (Figure 2A). The terminal stem-loop structure of FsoOLV1 was predicted using the UNAfold web server. The results showed that the first 33 nucleotides at the 5′ terminus could form a stable stem-loop structure, while the final 39 nucleotides at the 3′ terminus formed two stable stem-loop structures (Figure 2B), which is a distinctive feature of ourmia-like viruses.

Homology searches using BLASTP revealed that the protein encoded by the ORF shares a conserved domain closely related to the RdRps of Plasmopara viticola lesion-associated ourmia-like virus 13 (YP_010800213) and Plasmopara viticola lesion-associated ourmia-like virus 19 (YP_010800219), with amino acid identities of 49.03% and 43.77%, respectively. Multiple sequence alignments performed using GeneDoc software indicated that the RdRp of FsoOLV1 contains seven conserved motifs, which further clarifies its genetic structure and relationships with other ourmia-like viruses (Figure 2C).

To investigate the phylogenetic relationship between FsoOLV1 and other mycoviruses, a phylogenetic tree was constructed using the maximum likelihood method (ML) based on the RdRp amino acid sequences of FsoOLV1 and related viruses in the Botourmiaviridae family, including Betascleroulivirus, Deltascleroulivirus, Gammascleroulivirus, Epsilonscleroulivirus, Ourmiavirus, Magoulivirus, Betabotoulivirus, Scleroulivirus, Betarhizoulivirus, Botoulivirus, Penoulivirus, and Rhizoulivirus, as well as viruses from the Narnarviridae family. FsoOLV1 shares a close phylogenetic relationship with the genus Magoulivirus and was grouped in the same clade as Plasmopara viticola lesion-associated ourmia-like virus 13 (Figure 3). Based on these results, FsoOLV1 was classified as a member of the genus Magoulivirus within the family Botourmiaviridae.

Pathogenicity testing of the FsoOLV1-containing strain SJH 2-4 and virus-free strain TH 7-2 on Panax ginseng roots revealed that FsoOLV1 reduced the virulence of F. solani, as previously described (Figure 1C). To further investigate the effects of FsoOLV1 on the growth and pathogenicity of F. solani, isogenic FsoOLV1-free strains, designated 2-4 VF 1-7, were derived from SJH 2-4 through protoplast regeneration (Supplementary Figure S2). Three of these virus-free strains, 2-4 VF1, 2-4 VF2, and 2-4 VF3, were selected for further study. When cultured on PDA medium for 7 days, the virus-free strains exhibited a faster growth rate compared with SJH 2-4 (Figures 4A, E). The hyphae of SJH 2-4 were irregular, slender, curved, and short-branched when cultured on SNA medium for 10 days, whereas the hyphae of the virus-free strains 2-4 VF1, 2-4 VF2, and 2-4 VF3 were neat and regular (Figure 4B). Additionally, SJH 2-4 has abnormally large vacuoles in its hyphal cells on SNA medium, a feature not observed in the three virus-free strains (Figure 4C). Similarly, after 5 days in CMC medium, SJH 2-4 exhibited malformed and shrunken macroconidia that were filled with numerous vacuoles, while the macroconidia of the virus-free strains appeared regular and healthy (Figure 4D). Spore production and mycelial biomass measurements showed that the virus-free strains 2-4 VF1, 2-4 VF2, and 2-4 VF3 had significantly higher conidial yields and mycelial biomass compared with the original SJH 2-4 strain (Figures 4F, H; Supplementary Table S2). The spore germination rate of the three virus-free strains ranged from 77.47% to 86.65%, which was notably higher than the 59.71% observed for strain SJH 2-4 (Figure 4G; Supplementary Table S2). Collectively, these results suggest that FsoOLV1 significantly affects the fungal host by altering the hyphal growth rate, mycelial biomass, hyphal and spore morphology, and spore germination rate.

To evaluate the effect of FsoOLV1 on the virulence of its host fungus, SJH 2-4 and the virus-free strains were tested on Panax ginseng roots. Seven days after fungal inoculation, the three virus-free strains led to severe root rot, while SJH 2-4 only resulted in small disease lesions (Figure 5A). The lesion areas caused by the three virus-free strains were notably larger than those caused by SJH 2-4 (Figure 5B). These observations indicate that FsoOLV1 is responsible for the hypovirulence of F. solani strain SJH 2-4.

To assess the horizontal transmission capabilities of FsoOLV1, dual-culture assays were conducted on PDA medium using the FsoOLV1-infected donor strain SJH 2-4 and two virus-free recipient strains, 2-4 VF1 and TH 7-2, respectively. After 5 days of co-cultivation, mycelial plugs were taken from three distinct regions on recipient sides and transferred to fresh PDA plates, yielding in six derived strains: VI VF1, VI VF2, VI VF3, VI TH-1, VI TH-2, and VI TH-3 (Figures 6A, B). RT-PCR analysis confirmed successful horizontal transmission of FsoOLV1 from SJH 2-4 to both recipient strains 2-4 VF1 and TH 7-2 (Figure 6C). However, efforts to transfer FsoOLV1 to other Fusarium species, including F. oxysporum, F. proliferatum, and F. verticillioides, were unsuccessful (data not shown).

To determine whether FsoOLV1 can be vertically transmitted, 21 single-conidium-derived colonies from strain SJH 2-4 were analyzed using RT-PCR with FsoOLV1-specific primers. The results showed that all isolates were FsoOLV1 positive (Figure 6D). Furthermore, no significant differences in growth rates were observed among these isolates when cultured on PDA plates. These findings indicate that FsoOLV1 is stably and efficiently transmitted through F. solani conidia.

To assess the infectivity of FsoOLV1 in other Fusarium species responsible for Panax ginseng root rot, FsoOLV1 was transfected to F. oxysporum, F. proliferatum, and F. verticillioides via protoplast transformation. Transformants were obtained from F. oxysporum strain FS 7-7-3 (FoVI1, FoVI2, FoVI3), F. proliferatum strains DH 8-9 (FpVI1, FpVI2, FpVI3), and F. verticillioides strain BS 1-7 (FvVI1, FvVI2, FvVI3), using PEG-mediated protoplast transfection. The control strains, FsVI1, FsVI2, and FsVI3, were derived from F. solani strain TH 7-2. RT-PCR detection confirmed the successful transfection of FsoOLV1 into the four Fusarium species (Figure 7A). Compared with the original strains, the FsoOLV1 transformants had slower growth rates, reduced spore production, lower spore germination rates, and decreased mycelial biomass (Figure 7B; Supplementary Figure S3; Supplementary Table S3). Microscopic observation revealed that the hyphae of FsoOLV1 transformants were twisted, sparse, and less branched (Supplementary Figure S4). Additionally, numerous vacuoles were observed in the hyphal cells of the FsoOLV1 transformants (Supplementary Figure S5). When Panax ginseng roots were inoculated with the original strains, they developed significantly more severe root rot symptoms compared with those inoculated with the FsoOLV1-carrying transformants (Figures 8A, B). These findings indicate that FsoOLV1 can confer hypovirulence in other Fusarium species. Based on these results, FsoOLV1 could be used as a biocontrol agent for managing ginseng root rot diseases.