Cnidium officinale Makino (family Umbelliferae), known as “Cheongung” in Korea, “Senkyu” in Japan, and “Chuanxiong” in China, is an economically important and culturally significant medicinal herb widely cultivated in East Asia (Kim et al., 2020). Its rhizomes exhibit anti-inflammatory, analgesic, and antioxidant activities, which have made C. officinale a staple in traditional medicine (Bae et al., 2011). These therapeutic effects are attributed to high levels of bioactive compounds such as phthalides, coumarins, and essential oils (Oh et al., 2010). Because C. officinale is propagated vegetatively, viruses can accumulate with each planting cycle, threatening both rhizome yield and the levels of active compounds (Bae et al., 2011). To date, virus-free propagation programs (e.g., clean seed-rhizome production) are lacking, in part because the full spectrum of viruses infecting this crop has not yet been fully characterized.

Despite its importance, the virome of C. officinale remains poorly characterized, with only a few viruses reported to date. For example, studies in Korea identified cnidium vein yellowing virus-1 (CnVYV-1) and -2 (CnVYV-2) associated with vein-yellowing symptoms (Yoo et al., 2015). A recent HTS-based analysis significantly expanded our knowledge by identifying additional viruses, but it focused on a single region and lacked detailed characterization of genome structures and regional distribution (Belete et al., 2024). Comprehensive virome characterization using HTS can provide critical insights into the diversity, evolutionary dynamics, and epidemiology of viruses infecting medicinal crops (Roossinck et al., 2015; Massart et al., 2017). Accurate identification of virus diversity and distribution is crucial for developing effective management strategies. This is especially true for medicinal crops like C. officinale, whose pharmacological quality can be directly impacted by viral infections – in particular, the timing of infection (Jeong et al., 2014; Fraile and García-Arenal, 2016; Mehetre et al., 2021).

Complex virus co-infections in plants can significantly alter host–virus interactions, symptom expression, transmission dynamics, and overall pathogenicity, influencing crop productivity and quality (Hadidi et al., 2016). HTS enables comprehensive, unbiased detection of both known and novel viruses without prior genomic information, making it particularly suitable for virome studies (Hadidi et al., 2016, 2012; Barba et al., 2014; Roossinck et al., 2015; Wu et al., 2015; Maree et al., 2018; Villamor et al., 2019; Maclot et al., 2020; Kutnjak et al., 2021; Maina et al., 2024). Modern plant viromics has revealed extensive viral diversity across hosts and environments and has accelerated discovery and diagnostics (Roossinck et al., 2015; Minicka et al., 2020). In practice, HTS can be implemented using rRNA-depleted total RNA, dsRNA or VANA enrichments, small-RNA sequencing, or, where appropriate, long−read platforms for rapid or field-deployable analyses; target−enrichment (hybrid-capture) can further improve sensitivity for low−titer viruses (Pooggin, 2018; Boykin et al., 2019; Sun et al., 2022; Briese et al., 2015; Farrall et al., 2024). Adoption in plant-health programs is supported by international guidelines that address validation, controls, and reporting (Massart et al., 2022).

In this study, we utilized HTS to systematically investigate the virome of symptomatic C. officinale plants from nine distinct geographic regions of Korea. Our specific objectives were: (i) to catalog the known and novel viruses infecting C. officinale; (ii) to quantitatively assess the relative abundance (RPKM) and distribution of identified viruses across different cultivation regions; and (iii) to elucidate the evolutionary relationships, genetic diversity, and complex co-infection patterns (e.g., antagonism and co-occurrence) of predominant viruses – particularly ASGV and CMV – through detailed phylogenetic and population genetic analyses. These findings provide valuable baseline data for further research into virus epidemiology and host–virus interactions. Moreover, they inform effective disease management strategies, ultimately enhancing the sustainable production and medicinal quality of C. officinale.

In this study, we conducted a comprehensive HTS survey of viruses infecting C. officinale across nine major cultivation regions in Korea. This represents the first broad virome characterization for this medicinal, vegetatively propagated crop. We identified 31 distinct viruses, including several previously known plant viruses and four novel species (provisionally named CnoV1, CnVY, CnVZ, and the satellite RNA satCnV). The application of HTS enabled not only the discovery and cataloging of these viruses, but also an analysis of their relative abundances and co-infection patterns across different regions. These findings significantly expand our understanding of virus diversity in C. officinale and provide new insights into virus community ecology in clonally propagated medicinal crops.

Notably, the dominant viruses infecting C. officinale – including CMV, CnVYV (a vein-yellowing virus complex with two species), ASGV, and CnVX – exhibited remarkably high genetic uniformity across all surveyed regions. Sequence variants from geographically distant provinces were often nearly identical, and phylogenetic analyses consistently clustered isolates by host species (i.e., C. officinale) rather than by location. For example, CMV and ASGV sequences from C. officinale formed host-specific clades distinct from isolates of the same viruses infecting other host plants (Figures 5-7). The extremely low genetic diversity observed for viruses like CMV and ASGV in C. officinale suggests a long-term, closed association with this host. All sampled isolates of these viruses were nearly identical across widely separated provinces. They were also highly similar to one another but distinct from strains found in other crops in Korea. This implies very limited gene flow between C. officinale virus populations and those in other host species. Once a virus was introduced into C. officinale, it likely became genetically isolated and was maintained almost exclusively within C. officinale through vegetative propagation. Such host-restricted virus lineages have been noted in other systems and are thought to arise when viruses persist in a crop over long periods without frequent reintroduction from outside sources (Kawakubo et al., 2023). Our data therefore point to a pattern of long-term, stable maintenance of viruses in C. officinale with minimal diversification, most likely as a consequence of the plant’s clonal propagation habit (which limits opportunities for new viral genotypes to be introduced or selected). This situation contrasts with seed-propagated or more widely exchanged crops, where viral populations often show greater regional genetic differentiation. For instance, a virome study of sweet potato (another vegetatively propagated crop) found that while multiple viruses are ubiquitous, many virus isolates (especially potyviruses) exhibited grouping by geographical region (Jo et al., 2020). In C. officinale, by contrast, we did not observe region-specific virus clades – an indication that C. officinale viruses have been circulating largely within a closed host population, with very limited new introductions or host shifts.

Beyond the major viruses, our survey uncovered additional viral elements that could influence disease outcomes. A particularly noteworthy discovery is a novel satellite RNA, designated satCnV, associated with the C. officinale virome. To our knowledge, this is the first report of a satellite RNA in viruses infecting C. officinale. Satellite RNAs are small sub-viral molecules that depend on a helper virus for replication and encapsidation. They are known to modulate disease severity and viral accumulation in various plant–virus systems (Palukaitis, 1988; Collmer and Howell, 1992; Hu et al., 2009; Cao et al., 2019). For example, satellite RNAs in other pathosystems (e.g., those associated with cucumber mosaic virus or bamboo mosaic virus) can attenuate or exacerbate symptom expression in the host (Lin and Lin, 2017). The discovery of satCnV in C. officinale raises intriguing questions about its biological role: it may interact with its helper virus (likely the CnVYV complex, as discussed below) to influence symptom development or virus multiplication in this crop. Determining whether satCnV ameliorates or aggravates disease symptoms, and how it affects the epidemiology of its helper virus, will be an important subject for future research.

Our analysis of co-infection patterns revealed both antagonistic and synergistic relationships among the C. officinale viruses. In particular, we observed a strong inverse correlation between the abundance of CnVY (the single-segment vein-yellowing virus) and CMV across the pooled regional samples. At the regional level, libraries where CMV was extremely abundant tended to have lower representation of CnVY, and vice versa. For instance, CMV dominated the virome in Bonghwa and Geochang, whereas CnVY was prevalent in regions like Samcheok-B and Jecheon. It is important to note that this negative correlation reflects opposite abundance trends in pooled samples, rather than an absolute mutual exclusion within individual plants. In fact, both viruses were detected (often together) in all regions, and in some areas (e.g., Hoengseong and Pyeongchang) they were simultaneously abundant. Thus, their relationship appears to be a broad ecological antagonism rather than a strict within-plant incompatibility. One plausible explanation is competitive interactions or cross-protection: infection by one virus might suppress superinfection by the other, especially if one induces a strong RNA silencing response or occupies the available cellular “niche” in the host (Hanssen and Thomma, 2010; Folimonova, 2012; Boezen et al., 2023). Cross-protection is a known phenomenon where a plant infected by a mild strain of a virus is protected against a more severe strain of a related virus (Mascia and Gallitelli, 2016). In our case, if CMV and CnVY induce cross-reactive defense responses or compete for the same host resources, an increase in one could broadly limit the other. Another contributing factor could be differences in their insect vectors or transmission efficiency: if management or environmental factors in certain regions suppress the vector of one virus, the other virus (with a different vector) might gain a relative advantage. In any event, the regional-scale antagonism between CMV and CnVY cautions that controlling one virus in the field could inadvertently create ecological space for the other to proliferate. This interplay underscores the complexity of virus–virus interactions in mixed infections, which can significantly impact disease outcomes (Syller, 2012; Mascia and Gallitelli, 2016).

In contrast to the CMV–CnVY inverse relationship, we found a tightly linked cluster of positive correlations among the CnVYV-1 and CnVYV-2 segments (the two components of the vein-yellowing virus complex) and the satellite satCnV. Unsurprisingly, the two genomic segments of CnVYV co-occurred consistently, as both are required for a successful infection by this bipartite virus. More interestingly, the satellite RNA was almost always present alongside CnVYV, with read counts of satCnV closely tracking those of CnVYV-1 and CnVYV-2 in our regional pools. This suggests that satCnV is dependent on CnVYV as its helper virus and may have a functional association with that viral complex. The strong tripartite correlation hints that we are observing a stable satellite–helper virus partnership in C. officinale. It will be fascinating to investigate how satCnV influences the biology of CnVYV infections. In other well-studied systems, satellites can attenuate the damage of their helper viruses or, conversely, exacerbate symptoms, depending on the specifics of their interaction (Palukaitis, 1988; Collmer and Howell, 1992; Hu et al., 2009; Roossinck, 2015). For CnVYV and satCnV, experiments under controlled conditions (e.g., co-inoculating plants with CnVYV with and without the satellite) could elucidate whether satCnV mitigates symptom severity or perhaps impacts the transmission or persistence of the virus. The discovery of this satellite therefore opens a new line of inquiry into the pathogenicity and epidemiology of the vein-yellowing disease in C. officinale.

A striking aspect of our findings is the minimal regional differentiation in virus composition and sequence variants, which we attribute to C. officinale’s propagation practices. Because this crop is propagated almost exclusively by dividing rhizomes (clonal propagation), viruses are readily passed from mother plants to all vegetative progeny. Over years of cultivation, a region’s virome is likely shaped by whatever viruses were present in the initial planting stock (seed rhizomes) introduced to local farms, followed by amplification of those viruses through clonal spread within that region. In essence, each region’s virus community may represent a founder effect of the viruses present in the source propagative material, with relatively little influx of new viruses unless growers exchange plant material across regions (Delatte et al., 2007; Sun et al., 2013; Razavi et al., 2025). Consistent with this, our data show that the same few viruses (CMV, CnVYV, CnVY, CnVX, ASGV, etc.) dominate across all areas, and their sequences are nearly identical countrywide. This pattern is best explained by propagation-driven epidemiology rather than independent viral invasions in each locale. If farmers in different provinces originally obtained their C. officinale rhizomes from one or a few common nurseries (or from each other), those source plants would have “seeded” the observed regional viromes. Thereafter, local vegetative multiplication of infected stock would ensure the perpetuation of the same viruses in each clonal lineage. The strong host-specific clade formation discussed above is a natural consequence of this closed transmission loop. Indeed, vegetative propagation is known to facilitate long-term virus maintenance and can lead to an accumulation of viruses in clonally propagated crops. Sweet potato is an illustrative example: nearly all sweet potato plants in the field carry multiple viruses, largely because cuttings from infected mother plants disseminate viruses perpetually, and the lack of regular seed propagation prevents virome reset. Similarly, in C. officinale, the vegetative spread likely allows viruses to persist indefinitely once established, creating a relatively static virome within the crop.

Alternative factors such as vector-mediated transmission or differences in crop management could also influence regional virus prevalence, but we suspect these play a secondary role in this case. C. officinale is not known for extensive long-distance vector spread of its viruses (unlike, say, sweet potato where whitefly and aphids contribute to virus dissemination (Jo et al., 2020)). The near-uniformity of virus sequences across distant farms suggests that recurrent reintroductions or active vector-driven mixing between regions have been very limited. If vectors were regularly moving viruses between C. officinale fields or from other host species into C. officinale, we would expect more genetic variation or region-specific strains, which we did not observe. Therefore, we infer that clonal propagation is the dominant force structuring the virome of this crop, in line with what has been proposed for other vegetatively propagated plants (Gibbs et al., 2017; Jones and Naidu, 2019). Our conclusions on propagation-driven virome structure are necessarily somewhat indirect because our sampling did not include tracing of planting material sources. Future studies should explicitly examine the infection status of C. officinale propagation stocks—such as nursery rhizomes or “seed” bulbs—and track their distribution routes. If, for instance, a given nursery’s rhizomes are found to carry a specific set of viruses, and those viruses match the profiles in client farms across a region, it would strongly corroborate the idea that the regional viromes were seeded via that infected planting stock. Additionally, surveying potential insect vectors in these regions and testing whether they carry C. officinale viruses could help clarify the role of horizontal (vector-borne) spread. However, given our current data, the most parsimonious explanation is that the virome population structure is largely clonal, reflecting historical propagation practices with limited inter-regional exchange of infected plant material.

The findings of this virome study have direct practical implications for improving the health and quality of C. officinale crops. We identified a suite of persistent viruses that can infect C. officinale, many of which (like CMV, CnVYV, ASGV, etc.) are capable of causing latent or chronic infections. In a vegetatively propagated host, an infection is essentially permanent: once a mother plant is virus-infected, all of its clonal offspring will carry that virus load. This highlights the critical importance of establishing virus-free propagation programs. Similar “clean seed” or clean stock initiatives in other crops have been highly successful in curbing virus diseases – for example, certified virus-free seed potato programs and sweet potato virus indexing schemes have dramatically reduced yield losses caused by viruses (Gergerich et al., 2015). HTS-based virome analysis, as we have demonstrated, can guide the development of such programs by revealing which viruses are prevalent and must be targeted. In fact, there is a growing recognition that virome indexing should be integrated into plant biosecurity and seed systems, especially for vegetatively propagated crops (Alcalá Briseño et al., 2023). Clean germplasm production relies on sensitive detection of viruses in source plants, and our comprehensive list of C. officinale viruses now enables the design of molecular diagnostic tools (such as RT-PCR or RT-LAMP assays) for each major pathogen. We therefore urge the implementation of a virus-free certification program in C. officinale cultivation. In such a program, candidate mother stock plants would be rigorously tested for the presence of the key viruses identified in this study, and only those testing negative (virus-free) would be used to produce new rhizomes for distribution. By continuously indexing propagation material for viruses, growers can break the cycle of passing along infections and gradually purge the regional viromes from their fields.

To make virus indexing practical, we propose a two-tier diagnostic panel based on the viruses we found to be most widespread or economically significant. Tier 1 (core panel) would include the highest-priority targets for routine screening: CMV, CnVYV-1 and -2 (vein-yellowing virus segments), CnVY, CnVX, ASGV, Cnidium polerovirus 1 (CnPV1), and the satellite RNA satCnV. These agents were either ubiquitous across regions or associated with notable disease symptoms, making them crucial to detect and eliminate. Tier 2 (supplemental panel) would encompass additional viruses that might be included based on regional context or specific epidemiological concerns. This could include the novel viruses CnoV1, CnV1, CnV2, CnVZ (which were found in more limited instances), as well as other localized viruses such as Angelica bushy stunt virus (AnBSV) if relevant. In practice, a multiplex RT-PCR or a similar high-throughput assay could be developed to simultaneously test for all Tier 1 viruses in a single reaction per sample – an approach that many plant certification programs have employed successfully. Tier 2 agents can be tested in a follow-up manner or included in expanded panels when tracing back the source of an infection. By prioritizing the most common and damaging viruses for regular indexing, the workload remains feasible while still covering the majority of infection risks.

In implementing these recommendations, coordination among growers, nurseries, and plant protection authorities will be key. Growers should be educated about the long-term benefits of planting virus-free C. officinale rhizomes, even if that requires discarding some existing infected stocks. Meanwhile, nurseries or tissue-culture labs could be enlisted to maintain clean mother blocks of C. officinale – collections of plants that are tested and confirmed free of the Tier 1 viruses – from which propagation materials are multiplied under protected conditions (e.g., insect-proof greenhouses). Such a system would mirror the clean plant programs in horticultural industries like fruit trees and sweet potatoes. Over time, the availability of certified virus-free rhizomes would allow farmers to replace their diseased planting material, leading to healthier crops with higher yields and more consistent medicinal quality. Our virome dataset provides the foundational knowledge to initiate this process, as we now know exactly which pathogens to screen for. In a broader sense, this study exemplifies how HTS-driven virome analysis can be translated into practical strategies for crop improvement. By revealing the hidden burden of viruses in clonally propagated crops and pinpointing the culprits, we can design informed interventions to break the cycle of infection (Alcalá Briseño et al., 2023).