Characterization of the phyllosphere virome of fresh vegetables and potential transfer to the human gut

Park, Ji-Woo; Yun, Yeo-Eun; Cho, Jin Ah; Yoon, Su-In; In, Su-A; Park, Eun-Jin; Kim, Min-Soo

doi:10.1038/s41467-025-58856-6

Download PDF

Article
Open access
Published: 10 April 2025

Characterization of the phyllosphere virome of fresh vegetables and potential transfer to the human gut

Nature Communications volume 16, Article number: 3427 (2025) Cite this article

6760 Accesses
1 Citations
2 Altmetric
Metrics details

Subjects

Abstract

Fresh vegetables harbor diverse microorganisms on leaf surfaces, yet their viral communities remain unexplored. We investigate the diversity and ecology of phyllosphere viromes of six leafy green vegetables using virus-like particle (VLP) enrichment and shotgun metagenome sequencing. On average, 9.2 × 10⁷ viruses are present per gram of leaf tissue. The majority (93.1 ± 6.2%) of these viruses are taxonomically unclassified. Virome compositions are distinct among vegetable types and exhibit temporal variations. Virulent phages with replication-enhancing auxiliary metabolic genes (AMGs) are more dominant than temperate phages with host fitness-benefiting AMGs. Analysis of 1498 human fecal VLP metagenomes reveals that approximately 10% of vegetable viruses are present in the human gut virome, including viruses commonly observed in multiple studies. These gut-associated vegetable viruses are enriched with short-term vegetable intake, and depleted in individuals with metabolic and immunologic disorders. Overall, this study elucidates the ecological contribution of the fresh vegetable virome to human gut virome diversity.

The gut virome in two indigenous populations from Malaysia

Article Open access 03 February 2022

Gut virome profiling identifies a widespread bacteriophage family associated with metabolic syndrome

Article Open access 23 June 2022

Interpersonal variability of the human gut virome confounds disease signal detection in IBD

Article Open access 25 February 2023

Introduction

The westernization of modern societies has brought about a decline in the biodiversity of the human gut microbiome^1,2, in part because of reduced exposure to the natural environment^3,4. Dietary plant fibers significantly influence the diversity and activity of the gut microbiome as these fibers serve as carbon and energy sources for fermentable bacteria⁵. Accordingly, frequent intake of plant-based foods helps to maintain the diversity of the gut microbiome in industrialized countries⁶. However, current research primarily emphasizes the impact of the dietary fiber content of these foods, overlooking the potential contribution of the microorganisms they carry.

Fresh vegetables harbor different types of microbial populations on their leaves⁷. The diversity of the vegetable phyllosphere microbiome is influenced by cultivation region^8,9 and season¹⁰, agricultural practices⁷, and postharvest handling^11,12. The phyllosphere microbiome improves crop quality during the preharvest stage¹³, while also influencing crop safety during the postharvest stage from production facilities to consumers¹⁴. Similar to microorganisms found in fermented foods^15,16, those present on fresh vegetables can be transferred to the gut¹⁷ and may therefore influence human health. Although the importance of the vegetable microbiome is recognized, most research has focused on bacterial and fungal communities within vegetables.

Viruses, predominantly bacteriophages, are key members of microbial communities in natural ecosystems¹⁸. Viruses shape bacterial community composition through their lytic lifestyle¹⁹ and modulate the functional phenotype of bacterial hosts via auxiliary metabolic genes (AMGs)²⁰. They also indirectly influence eukaryotic hosts through microbiome alterations^21,22. Despite their ecological significance, characterizing viral communities (viromes) on fresh vegetables remains challenging, primarily because of difficulties in directly sequencing limited amounts of viral DNA²³. Moreover, limitations in annotating viral sequences hinder understanding of the dynamics and functions of viromes, particularly in the human gut²⁴. Therefore, investigating the fresh vegetable virome and its potential transmission to the gut is expected to elucidate the ecological origins of gut-resident viruses and their potential impact on health.

In this study, we investigated the phyllosphere virome of six leafy green vegetables (broccoli, spinach, perilla, red lettuce, green lettuce, and romaine lettuce) using a combination of virus-like particle (VLP) enrichment and shotgun metagenome deep-sequencing without prior random amplification. We characterized the total viral load, and the diversity, variability, and functional potential of the vegetable phyllosphere virome. Based on our de novo vegetable virome database, we explored the potential transfer of vegetable-associated viruses to the human gut by comparing their genomes with 1498 fecal VLP metagenomes from 13 studies (huVLPs) and those in human gut virome databases (GVDs). The results of this study enhance our understanding of the viral ecology of fresh vegetables, and illuminate the role of the food-gut axis in shaping and maintaining the healthy human gut virome.

Results

Reconstruction of viral genomes from the VLP metagenomes of fresh vegetables

Leaf samples of six vegetables (n = 60) were collected from retail stores between 2019 and 2021 (Supplementary Fig. 1 and Supplementary Data 1). Leaf samples were homogenized to detach epiphytic and endophytic microorganisms including viruses, and then, VLPs were isolated. The metagenomic DNA of VLPs was deep-sequenced without random amplification, which avoids composition bias toward single-stranded circular viral genomes^23,25. After removing plant-derived reads, an average (±s.d.) of 2,186,429 ± 3,724,274 high-quality paired-end reads per sample were obtained, accounting for 9.9 ± 15.0% of raw reads. Sixteen samples failed due to insufficient VLP-derived DNA for sequencing directly. The reads were assembled into 19,918 large contigs (≥2.5 kb). A combination of four in silico viral sequence identifications found 9773 viral contigs, accounting for 33.8 ± 33.0% of the high-quality reads (Supplementary Fig. 2a). The completeness of viral contigs was estimated using CheckV²⁶. One-fourth (23.8 ± 25.7%) of viral contigs had no amino acid similarities with previously known genomes. The remaining contigs were classified into low-quality (66.5 ± 23.0%), medium-quality (5.2 ± 6.5%), high-quality (2.9 ± 4.0%), and complete viral genomes (1.7 ± 2.5%). In total, the qualified contigs accounted for 78.3 ± 26.8% of the reads mapped to the viral contigs (Supplementary Fig. 2b), suggesting that a substantial number of vegetable viruses are unexplored. Finally, a virome dataset comprising 8137 non-redundant viral genomes was established, which consisted of 198 circular (40,201 ± 28,855 bp) and 7939 linear genomes (9060 ± 13,088 bp). Using ViromeQC²⁷, we quantified prokaryotic ribosomal RNA and single-copy marker genes in five datasets, each representing different stages of sequence processing: raw reads, preprocessed reads, high-quality reads, reads for large contigs, and reads for viral contigs. The viral enrichment score sharply increased from raw reads to reads for viral contigs (P < 0.0001) (Supplementary Fig. 2c), indicating that this study successfully obtained genomic sequences enriched in viral information from fresh vegetable samples.

Fresh vegetables harbor an abundant and specific phyllosphere virome

We quantified the total viral abundance of vegetable viromes by adding Escherichia coli T4 phage to samples, the amount of which was pre-determined by epifluorescence microscopy²⁸. Samples treated with T4 phage had an average of 2587 ± 6446 reads (0.2 ± 0.8%), whereas untreated samples had no T4 phage reads (P = 0.0003) (Supplementary Fig. 2d), validating the use of T4 phage as an internal standard for quantifying the total viral load of fresh vegetables. Based on the length-normalized read number of the T4 phage genome, an average of 9.8 × 10⁷ ± 2.2 × 10⁸ viral genomes were estimated per gram of leaf. The number of viral genomes was highest in spinach, lowest in perilla, and similar among the others (P < 0.05) (Fig. 1a). The number of viral genomes was significantly lower in pesticide-free samples than in conventional and organic samples (P < 0.05) (Fig. 1b). The number of viral genomes did not differ between sampling periods or distributors (Fig. 1c, d). These results suggest that numerous viruses are present on fresh vegetables, and that their abundance varies depending on vegetable type and farming method used. We also quantified total bacterial abundance and estimated an average of 3.1 × 10⁷ ± 1.3 × 10⁸ bacterial genomes per gram of leaf. Similar to viral genomes, the number of bacterial genomes was highest in spinach (Fig. 1e), and lowest in pesticide-free samples (P < 0.01) (Fig. 1f). The number of bacterial genomes did not differ between sampling periods or distributors (Fig. 1g, h). These similarities between the total viral and bacterial abundances suggest that the two are closely linked. Based on these observations, an average of 1.2 ± 12.7 virus-to-bacteria ratio (VBR) was determined across vegetables (Supplementary Fig. 2e). According to recent metagenome-based VBR estimations²⁹, these values fall within the expected range for planktonic ecosystems, although the VBRs were variable. Considering that VBR roughly serves as an indicator of viral infection strategies in a given environment³⁰, these results suggest that viral infection dynamics in the phyllosphere of fresh vegetables may be similar to those observed in relatively nutrient-poor environments³¹, wherein lytic dynamics dominate.

**Fig. 1: Variations in the total abundance, diversity, and composition of vegetable viromes and bacterial microbiomes by vegetable type.**

The viral genomes were clustered into species-level taxonomic groups³², resulting in a total of 6363 virus operational taxonomic units (vOTUs; 239 ± 485 vOTUs) (Supplementary Data 2). Significant differences were observed in the number of vOTUs and/or the Shannon index among the vegetable types and sampling periods (P < 0.05) (Fig. 1i, j), but not among the farming methods or distributors (Supplementary Fig. 3a, b). Virome composition was compared according to four host and environmental factors (vegetable type, sampling period, farming method, and distributor) using distance-based redundancy analysis (db-RDA). The best-fit model was constrained to vegetable type and sampling period, which explained 15.6% and 8.9% of total variance, respectively (P < 0.05) (Fig. 1k). Specifically, the viromes of the three lettuce types were separated from those of non-lettuces along the first axis of the db-RDA (P = 0.020), and the viromes of the three sampling periods were separated along the second axis (P = 0.001). These results suggest that the vegetable virome is host-specific, and changes temporally.

Similar patterns were observed in the bacterial microbiome. Based on 1154 amplicon sequence variants (ASVs; 116 ± 52 ASVs), significant differences were observed in the number of ASVs and the Shannon index among vegetable types (P < 0.05), but not among sampling periods, farming methods, or distributors (Fig. 1l, m and Supplementary Fig. 3c, d). The best db-RDA model fitted by vegetable type and sampling period was selected, which explained 21.8% and 3.3% of the total variance, respectively (P < 0.05) (Fig. 1n). The bacterial microbiomes of the three lettuces were distinct from those of non-lettuce vegetables along the first axis (P = 0.001), and the bacterial microbiomes of the two sampling periods were separated along the second axis (P = 0.004). Therefore, strong co-variation between viral and bacterial microbiomes suggests that the ecology of the virome is closely associated with the bacterial microbiome in the vegetable phyllosphere.

Fresh vegetable viruses are largely uncharacterized

We next determined the taxonomic diversity of the vegetable virome according to the International Committee on Taxonomy of Viruses taxonomy³³. A gene-sharing network was constructed with the viral genomes of the 2003 vOTUs (31.5%), the NCBI RefSeq Viral Genome, and IMG/VR using vConTACT2³⁴, resulting in a total of 1062 viral clusters (VCs; 65 ± 108 VCs) representing genus-level taxonomic groups (Fig. 2a). Of those, 107 vOTUs (5.3%) in 64 VCs were classified into 49 viral genera of nine families, and 433 vOTUs in 147 VCs were assigned to 10 families. As a result, 5823 vOTUs (93.1 ± 6.3%) could not be classified into any genus or family (Fig. 2b), underscoring that viruses present on fresh vegetables are largely uncharacterized.

**Fig. 2: Taxonomic diversity of fresh vegetable viromes.**

Taxonomically identified vOTUs were mostly classified as prokaryotic viruses (98.4 ± 3.6%) (Fig. 2b), with most assigned to four families within the Caudovirales order: Siphoviridae (50.4 ± 38.1%), Myoviridae (29.2 ± 30.1%), Podoviridae (6.9 ± 9.5%), and Autographiviridae (6.5 ± 9.7%) (Fig. 2c). Viral abundance was largest in the Siphoviridae family (44.8 ± 42.3%), followed by the Myoviridae (31.9 ± 38.0%) and Autographiviridae families (9.1 ± 15.6%) (Fig. 2c). Single-stranded Microviridae vOTUs were observed only in spinach (20.0 ± 44.7%) and green lettuce (24.6 ± 28.7%). At the genus-level, 31 genera (63.3%) were observed in only a single vegetable type, while Lambdavirus was observed across four vegetables (Fig. 2d). These results suggest that host-specific bacteriophages dominate the phyllosphere virome of fresh vegetables.

Virulent and temperate phages differently interact with their bacterial hosts

To understand how vegetable viruses interact with their hosts, we focused on viral lifestyles by searching for lysogeny-associated genes and host regions flanking viral genomes. A total of 484 vOTUs were identified as temperate phages (8.5 ± 7.4% of the vOTUs and 17.5 ± 23.7% of the total abundance) (Fig. 3a), indicating that virulent phages predominate the vegetable virome, consistent with the VBR prediction in Supplementary Fig. 2e. This fraction remained relatively consistent regardless of genome completeness and sampling periods (Supplementary Fig. 4a, b). Next, we predicted prokaryotic hosts of virulent and temperate vOTUs using iPHoP³⁵, combined with sequence matches of CRISPR spacers and tRNAs. A total 1059 vOTUs (16.6%) were linked to a single bacterial host. For both phage types, hosts most frequently belonged to Gammaproteobacteria (59.1 ± 32.5% of virulent vOTUs; 40.3 ± 34.0% of temperate vOTUs), followed by Actinomycetes (13.9 ± 22.8%; 20.0 ± 28.2%) and Alphaproteobacteria (8.4 ± 13.9%; 36.2 ± 37.8%) (Supplementary Fig. 4c, d). However, the bacterial phyla, families, and genera infected with virulent vOTUs were significantly more diverse than those infected by temperate vOTUs (P < 0.05) (Fig. 3b), although this pattern varied depending on bacterial taxa (Fig. 3c). These results suggest that on vegetables, virulent phages are more common and interact more broadly with bacterial hosts than temperate phages.

**Fig. 3: Viral lifestyle and AMG composition in the virome of fresh vegetables.**

Next, we identified AMGs encoded in the genomes of virulent and temperate vOTUs using DRAM-v³⁶ and VIBRANT³⁷. A total of 458 AMGs were found in 345 virulent vOTUs, and 66 in 56 temperate vOTUs (Supplementary Data 3). DRAM-v identified the phoH (phosphate starvation-inducible) gene as the most common (36 AMGs), followed by AMGs related to xenobiotics biodegradation and metabolism (17) and glycan metabolism (9). VIBRANT identified AMGs involved in the metabolism of cofactors and vitamins (190), carbohydrate metabolism (190), and glycan metabolism (54) as the most common. Notably, the overall AMG composition differed significantly between virulent and temperate vOTUs (P = 0.012) (Fig. 3d). Specifically, AMGs involved in lipid metabolism, biosynthesis of other secondary metabolites, cellular community, and unclassified signaling and cellular processes were exclusive to virulent vOTUs (Fig. 3d). In addition, AMGs related to the metabolism of cofactors and vitamins, xenobiotics biodegradation and metabolism, and folding, sorting and degradation were enriched in virulent vOTUs, while energy metabolism AMGs were enriched in temperate vOTUs (P < 0.05) (Supplementary Fig. 4e). These findings align with the observations of Luo et al.²⁰ describing that virulent viruses modulate translation efficiency for rapid replication, while temperate viruses adjust host metabolism to provide competitive advantages to their hosts. For example, virulent vOTU1453 encodes queC, queD, and queE genes for the synthesis of 7-cyano-7-deazaguanine, potentially influencing translation efficiency³⁸ (Fig. 3e). By contrast, temperate vOTU1313 encodes cysH and mec genes, involved in assimilatory sulfate reduction³⁹ (Fig. 3f). Collectively, these results indicate that virulent and temperate viruses interact differently with their hosts in the microbiome of fresh vegetables.

Fresh vegetable viruses are transferred to the human gut

Based on environmental source information provided by IMG/VR, we found that the vegetable virome (927 vOTUs in 552 VCs) was associated with various ecosystems, including plant- (25.0 ± 13.7%) and soil-associated (22.7 ± 18.0%) ecosystems (Supplementary Fig. 5). Of these, we focused on 108 vOTUs (8.7 ± 19.0%) that were associated with viral genomes derived from human-associated ecosystem. Zhang et al.⁴⁰ demonstrated that the intestinal RNA virome largely consists of plant-infecting viruses derived from dietary plants, a finding corroborated by later studies^41,42. We hypothesized that similar observations could be made for the DNA gut virome. We compared vOTU genomes with those in IMG/VR and three GVDs (Gut Phage Database⁴³, human Gut Virome Database⁴⁴, and Metagenomic Gut Virus⁴⁵), and identified 61 vOTUs that were associated with 92 GVD genomes at the genus or species level (Supplementary Fig. 6a). Particularly, five vOTUs showed species-level genomic similarities with 24 GVD genomes (Supplementary Data 4), and vOTU16 had an identical genome to gvd30371 (Supplementary Fig. 6b).

Typically, only a small fraction of foodborne microorganisms is found in the human gut microbiome^17,41. GVDs are predominantly constructed from bulk metagenomes⁴⁶, which restricts in-depth viral detection, compared with VLP metagenomes^47,48. Based on these considerations, we tested our hypothesis using virome datasets generated through VLP enrichment, which was expected to allow more sensitive detection of vegetable-derived viruses. We collected 1498 VLP metagenomes from 1306 individuals across 13 studies, representing diverse demographics and health conditions (Fig. 4a and Supplementary Data 5). High-quality reads were compared with vOTU genomes. crAssphage genome was used as a control because of its high abundance and prevalence in the human gut virome⁴⁹. We identified 627 vOTUs (25 ± 40 vOTUs) as gut-associated (Fig. 4b and Supplementary Data 6). An average of 130 ± 97 vOTUs were detected per study, with low abundance (0.6 ± 1.6 fragments per kilobase million, FPKM, per vOTU and 41.4 ± 89.8 FPKM for total vOTUs) but high prevalence (97.8%), compared with crAssphage (109.6 ± 382.9 FPKM and 77.2%, respectively) (Fig. 4c). While this number appears to reach a plateau within huVLPs (Fig. 4d), it is expected to increase as additional studies are conducted (Fig. 4e). Of those, 101 vOTUs were observed in more than three studies, and 46 vOTUs were observed in more than four studies (Fig. 4c). These global vOTUs were derived from all vegetables, and were repeatedly observed in 1-year longitudinal samples⁵⁰ (Supplementary Fig. 7). For bacterial hosts, six global vOTUs were identified as Pseudomonas, four as Escherichia, three as Pedobacter, two as Chryseobacterium, and one each as Pantoea, Duganella, Burkholderia, Acinetobacter, and Erwinia. These genera have been reported in human gut metagenome-assembled genome datasets^51,52, and were recently identified as vegetable-derived intestinal bacteria¹⁷. However, these results were obtained from eight huVLP datasets, with only a few vOTUs (12 ± 7 vOTUs per study) detected in five of the datasets. To further investigate this, we classified huVLPs into three categories based on sequencing depth: high (≥3 million reads), medium (1–3 million reads), and low (<1 million reads). The five datasets with a significantly lower number of vOTUs belonged to the low- or medium-depth group (P < 0.05) (Fig. 4f), indicating the need for deep-sequencing to effectively detect vegetable viruses in VLP metagenomes. Collectively, these findings suggest that viruses from fresh vegetables are transferred to the human gut, albeit as a relatively small proportion of the gut virome.

**Fig. 4: Abundance and prevalence of gut-associated vegetable viruses in the human gut virome.**

Short-term vegetable intake induces viral transmission from fresh vegetables

We examined whether viral transmission from fresh vegetables occurs through the consumption of these vegetables. We recruited 53 healthy individuals (Supplementary Data 7) and analyzed their fecal virome based on short-term and long-term dietary patterns, characterized by intake amounts of 16 and 20 food groups, respectively. The composition of 66 gut-associated vOTUs varied significantly with the short-term intake of two food groups: vegetables and fruits (P < 0.05) (Fig. 5a). This was confirmed by redundancy analysis, which revealed that variations in the vOTUs were explained the most by four food groups in the following order: vegetables (r² = 0.0357), potatoes (r² = 0.0351), legumes (r² = 0.0245), and fruits (r² = 0.0230) (P < 0.05) (Fig. 5b). Specifically, 24 vOTUs showed significant enrichment or depletion with short-term intake of the four groups, with the largest positive associations observed for vegetable (P < 0.05) (Fig. 5c). Subjects with a high vegetable intake had significantly more vOTUs than those with low or medium consumption (P < 0.05) (Fig. 5b). These results provide evidence that frequent, short-term consumption of dietary vegetables may facilitate the transmission of viruses from fresh vegetables. These results were not replicated in long-term dietary patterns. This may also explain why the majority of vOTUs (49.2 ± 25.2%) had a prevalence of less than 10% among subjects in each study. However, the abundance of vOTUs in huVLPs did not correlate positively with their abundance and prevalence in the original samples (Supplementary Fig. 8a, b), suggesting that high vegetable consumption does not fully facilitate the equal transmission of all viruses to the gut. This is further reflected in the composition comparisons between bacterial genera in the vegetable virome and bacterial hosts in huVLPs. Indeed, only a subset of abundant vOTUs appears to be frequently transferred to the human gut (Fig. 5d). This cannot be explained by viral lifestyle as well (Supplementary Fig. 8c). These findings allow speculation that viral transmission from fresh vegetables may be a selective process.

**Fig. 5: Correlations between the composition of gut-associated vOTUs and short-term dietary patterns.**

Vegetable viruses are reduced in individuals with metabolic and inflammatory disorders

Disease-specific alterations in the intestinal virome (referred to as “virome dysbiosis”) have been observed in individuals with metabolic and inflammatory disorders^{24,53,54,55,56,57}. In this context, we characterized the dynamics of gut-associated vegetable viruses in virome dysbiosis associated with obesity⁵⁶, ulcerative colitis⁵⁴, colorectal cancer/adenoma⁵³, alcoholic liver disease (ALD)⁵⁵, and type 2 diabetes (T2D)⁵⁷. The composition of gut-associated vOTUs in the disease groups differed significantly from that in healthy controls across all conditions (P < 0.01) (Fig. 6a). Specifically, viral diversity (Shannon index) and/or relative abundance (FPKM) of the vOTUs was significantly lower in the disease groups than in healthy controls (P < 0.05) (Fig. 6a). Also, the abundance of most vOTUs was significantly lower in the disease groups than in healthy controls (P < 0.05) (Fig. 6b). These findings suggest that a reduction in vegetable-derived viruses is a common feature of the virome dysbiosis observed in metabolic and inflammatory disorders.

**Fig. 6: Alterations in the diversity and composition of gut-associated vegetable viruses in people with metabolic and inflammatory disorders.**

Discussion

The consumption of fresh vegetables benefits human health through multiple mechanisms, including the provision of bioactive compounds⁵⁸ and dietary fiber content, as well as the resulting microbial fermentation⁵⁹. While the potential influence of microorganisms from fresh vegetables would also be expected, evidence for the microbiome-food-gut axis has been largely restricted to lactic acid-producing bacteria in fermented foods^15,16. To explore this, the transmission of microorganisms from fresh vegetables to the host, and the interaction of microorganisms with the host’s gut microbiome requires elucidation. This study identified a high diversity, specificity, and variability of the phyllosphere viromes of fresh vegetables, as well as a virome-food-gut axis, revealing that fresh vegetables contribute significantly to the diversity and dynamics of the human gut virome.

The human gut virome is predominantly composed of DNA bacteriophages, with RNA viruses comprising a smaller proportion^42,50,60. While RNA gut viruses are known to be primarily plant pathogenic viruses introduced through the consumption of plant foods including vegetables⁴⁰, DNA gut viruses remain largely “viral dark matter” because of limited viral sequence information representing diverse environments, which hinders the identification of a healthy gut virome⁶¹. Our previous study suggested that temperate phages induced from lysogenic bacteria contribute to the composition of the DNA gut virome⁶², a finding confirmed by research on the infant gut virome⁶³ and gut-derived bacterial prophages⁶⁴. In this context, this study advances our understanding of human gut virome composition by demonstrating the horizontal virome transfer from fresh vegetables to the human gut. Specifically, on average, 10.0% of the vegetable virome was detected in global huVLP datasets, albeit at low abundance. This proportion is expected to increase as more viral genomes from fresh vegetables are analyzed with adequate sample size. Dozens of gut-associated vOTUs were shared across multiple, independent huVLP studies, each representing different regions, countries, and ethnicities. Although their abundance varied, their presence remained consistent within the same individuals over extended periods. Additionally, bacterial hosts associated with gut-associated vOTUs support the hypothesis of viral transmission from fresh vegetables, as these bacterial genera have been previously identified in human gut microbiomes^51,52. All these findings provide evidence that viruses from fresh vegetables reach the distal colon, which aligns well with the recent study of Wicaksono et al.¹⁷ demonstrating the presence of fruit and vegetable-derived bacteria in the human gut. Collectively, DNA viruses from fresh vegetables may be an integral component of the human gut virome, and fresh vegetables are an environmental reservoir for human gut DNA viruses, similar to the RNA gut virus⁴⁰.

Positive associations between gut-associated vOTUs and vegetable intake, both in terms of viral richness and composition, further support the hypothesis that viral transmission may occur through the consumption of fresh vegetables. These associations were observed only with short-term dietary patterns, rather than long-term patterns, suggesting that vegetable-associated viruses observed in the gut virome are likely acquired exogenously and transiently, rather than being part of a stable gut virome. This hypothesis is supported by high individuality in the abundance and prevalence of gut-associated vOTUs, which exhibited log-normal-like distributions in the huVLP datasets. Such distributions are similar to those observed for exogenously acquired microorganisms⁶⁵, providing additional evidence for their transient nature in the gut ecosystem. The findings of Carmody et al.⁶⁶ that raw plant foods disrupt gut microbiome physiology support this concept to be more feasible. Considering that most viral genera were specific to a particular vegetable type, consuming a variety of vegetables may contribute to a diverse individual gut virome⁶⁷. However, viral transmission from fresh vegetables may not operate equally for all viruses, which is supported by the lack of associations between the prevalence and abundance of vOTUs on vegetables and those in the human gut. For example, Pseudomonas, Escherichia, Chryseobacterium, and Pedobacter were identified as hosts for global vOTUs, and are all known to possess mechanisms that resist colonization resistance by forming biofilms^68,69,70 or interacting with the mucus layer⁷¹, allowing them to persist in the human gut. Moreover, the AMGs of gut-associated vOTUs, such as those encoding polygalacturonase for dietary pectin degradation, likely expand the metabolic niches of their hosts by enhancing nutrient utilization⁷². These characteristics suggest a model of viral transmission, wherein favorable viruses from fresh vegetables are selectively transmitted to the gut. Thus, transfer of viruses along with their bacterial hosts may be linked to the transmission and persistence of vegetable viruses in the gut.

Interestingly, a community of gut-associated vOTUs differed between individuals with metabolic and inflammatory disorders and healthy controls. These differences were characterized by a reduction in viral diversity and/or their relative abundance. Virome dysbiosis typically involves a reduction in overall viral diversity^{53,54,56,57,73}, driven by the induction of temperate Caudovirales phages⁷⁴. This suggests that the expansion of dysbiosis-associated phages may reduce the fraction of vegetable-derived viruses, potentially making them less detectable in sequencing data. Accordingly, our findings propose that gut-associated vegetable viruses could constitute a significant component of a healthy gut virome⁶¹ and may therefore be beneficial, rather than deleterious. While this behavior differs from the patterns observed in ectopic colonization by translocated oral bacteria in the intestine⁷⁵, the distinct behavior of vegetable-derived viruses in the gut virome could serve as a biomarker for the dysbiotic state of the gut virome. Intriguingly, not all datasets such as the ALD⁴² or T2D⁵⁷ studies exhibited reductions in viral diversity and abundance. Therefore, further validation of the “marker” hypothesis for gut-associated vegetable viruses is required through well-controlled experiments.

Based on our findings, consuming a typical serving of 85 grams of salad daily⁷⁶ results in the ingestion of ~7.8 billion viruses, a number that can be increased by prophage induction caused by sauces or sweeteners^77,78. Their significant presence in the human gut, non-random transmission patterns, and altered abundance in microbial dysbiosis suggest the hypothesis that prolonged exposure to these viruses could influence the composition and function of the gut microbiome. For example, Hsu et al.¹⁹ demonstrated that even a single foreign phage can disrupt the composition of the gut microbiome and its associated metabolites in a mouse model. While vegetable-derived viruses were present in low amounts and were short-lived in the gut, their ecological behavior appears similar to that of foodborne probiotics⁷⁹, suggesting that they might have comparable effects on the gut ecosystem, regardless of their abundance⁸⁰. Given that vegetable-derived viruses are integral to a healthy gut virome, they may play a role in modulating the microbiome and subsequent phenotypic changes in both human subjects with metabolic syndrome⁸¹ and chemical-induced colitis mice⁸², following fecal virome transplantation from healthy donors. These observations indicate that frequent exposure to vegetable-derived viruses could have transient, individualized impacts on the assembly and dynamics of the gut microbiome and, consequently, on human health.

Overall, our findings highlight the phyllosphere of vegetables as an overlooked part of the environmental virome and suggest that dietary vegetables act as vectors for the transmission of plant microbiome-associated bacteriophages to the human gut. Expanding the collection of viral genomes from a wider variety of fresh vegetables, capturing actively infecting or integrated viruses from bulk metagenomes, and addressing RNA viruses could enhance the detection of vegetable-originated viruses in the gut virome. Further experimental investigations, combined with strain-level analysis of gut viromes and paired vegetable viromes, are necessary to corroborate our findings. This study significantly advances our understanding of the viral ecology of fresh vegetables and the interactions of vegetable viruses with the human gut. Our findings may also serve as a basis for future studies to investigate the impact of food-derived microorganisms on human health through the microbiome-food-gut axis.

Limitations of this study

Although this study demonstrates that the phyllosphere virome of fresh vegetables is an understudied component of the plant microbiome, our de novo vegetable virome data represents only a fraction of the actual virome present on fresh vegetables. First, while the results of metagenomic analysis revealed high taxonomic diversity and variability in the phyllosphere virome, the limited sampling depth per vegetable type likely resulted in underrepresentation of the total viral diversity of fresh vegetables. Second, while fresh vegetables exist in two environmental states—preharvest and postharvest—our study deliberately chose to collect samples from retail stores for two strategic reasons: 1) this sampling approach reflects typical consumer access points, as most people obtain their vegetables from retail stores rather than directly from farms, and 2) it aligned with our primary research objective of investigating the potential dietary translocation of vegetable-associated viruses to the intestine. This is relevant since we typically consume fresh vegetables after they have undergone postharvest processing (harvesting, packaging, transportation, and storage). In this line of idea, our sampling approach focused on market-ready vegetables, thereby better representing the food microbiome rather than the host-associated microbiome of vegetable plants. This sampling approach captures the postharvest state of the phyllosphere virome, which may differ from its preharvest state. Third, despite collecting pre-packaged and securely sealed vegetable samples from the stores, potential contamination due to human handling at the store cannot be ruled out. To develop a comprehensive understanding of the phyllosphere virome, analyzing samples during the preharvest stage with sufficient biological replicates is required to capture spatial and temporal variation in virome composition and diversity.

Methods

Sample collection

Samples of broccoli (Brassica oleracea var. italica, n = 10), perilla (Perilla frutescens var. crispa, n = 9), spinach (Spinacia oleracea, n = 6), green leaf lettuce (Lactuca sativa, n = 11), red leaf lettuce (Lactuca sativa, n = 13), and romaine lettuce (Lactuca sativa, n = 11) were purchased from six retail stores located in Daejeon city and Pyeongtaek, Republic of Korea, during August 2019 to July 2021 (Supplementary Data 1). Samples labeled as conventional, pesticide-free, and organic were collected for different types of vegetables, according to their packaging labels. Samples were delivered and stored at 4 °C for no more than 3 days before further processing. Approximately 50–100 g of broccoli floret, 50–100 g of upper part of lettuce and perilla, and 10–30 g of a whole leaf of spinach were separated from the stem area using a sterile knife, and then cut into uniform pieces with a maximum width of 20 mm.

Preparation of Escherichia coli T4 phage

Escherichia coli T4 phage DSM 4505 was prepared, according to Bonilla et al.⁸³. Briefly, T4 phage was propagated by infecting E. coli DSM 613. Phage lysate was centrifuged at 4000 × g for 20 min at 4 °C, and the supernatant was filtered through 0.20-µm pore cellulose acetate syringe filters (Sartorius, Germany). The filtrate was mixed with 0.1 volume of chloroform (Sigma-Aldrich, USA) and incubated for 10 min at room temperature. The mixture was centrifuged at 4000 × g for 5 min at room temperature, and the supernatant was washed with 0.02-μm pre-filtered, sterilized sodium-magnesium (SM) buffer using AmiconⓇ Ultra-15 Centrifugal Filters (100 kDa MWCO, Sigma-Aldrich, USA). The phage concentrate was serially diluted, filtered through 0.02-µm Anodisc filters (Whatman, UK), and stained with 10× SYBR Gold (Invitrogen, USA). VLPs were then counted using an epifluorescence microscope (BX51-FL, Olympus, Japan). The phage concentrate was stored in a 40% glycerol solution at −70 °C. The stock solution was 10-fold diluted and used as an internal standard.

VLP enrichment, DNA extraction, and shotgun metagenome sequencing

Samples were homogenized with 100–300 ml of SM buffer in a sterile plastic bag (BagFilter®, Interscience, France) using a stomacher at speed 4 (BagMixer® 400 CC, Interscience, France) for 2 min. The homogenate was centrifuged at 8000 × g for 10 min at 4 °C, and the supernatant was transferred into a new tube. The pellet was stored at −70 °C for bacterial metagenome DNA extraction. This step was repeated once or twice to ensure complete collection of VLPs and microbial cells. A diluted solution of E. coli T4 phage was added to the supernatant in total amounts ranging from 5.0 × 10⁵ to 1.0 × 10⁸ VLPs (Supplementary Data 1). The supernatant was sequentially filtered through 0.45-µm and 0.2-µm cellulose acetate syringe filters (Sartorius, Germany). The filtrate was mixed with 10% (wt/vol) polyethylene glycol 8000 (Sigma-Aldrich, USA) and 1 M NaCl (Bioneer, South Korea), and incubated at 4 °C for 15 h. The mixture was centrifuged at 13,000 × g for 30 min at 4 °C, and the resulting pellet was resuspended in 5 ml SM buffer. The resuspension was further concentrated using 100 kDa MWCO centrifugal filters (AmiconⓇ Ultra-15 Sigma-Aldrich, USA). The concentrate was treated with 20 µl lysozyme (10 mg/ml, Biosolution, South Korea), and incubated at 37 °C for 30 min. It was further incubated with 0.2 volume of chloroform at 4 °C for 30 min, with vortexing every 5 min. Chloroform was removed by centrifugation at 4500 × g for 10 min at 4 °C. The concentrate was treated with three units of DNase I (Sigma-Aldrich, USA) and RNase A (Bioneer, South Korea) and incubated at 37 °C for up to 3 h to remove free nucleic acids. A stop solution was then added and incubated at 65 °C for 15 min. Total VLP DNA was extracted by treating 30 µl of 10% SDS (Sigma-Aldrich, USA) and 3 µl proteinase K (20 mg/ml, 600 mAU/ml, Bioneer, South Korea) at 55 °C for 1 hr. Subsequently, 100 µl of 5 M NaCl and 10% (v/v) cetyltrimethylammonium bromide buffer were added and incubated at 65 °C for 10 min. An equal volume of phenol/chloroform/isopropanol (25:24:1, Sigma-Aldrich, USA) was added, and the aqueous solution was collected by centrifugation at 8000 × g for 10 min at room temperature. Chloroform was repeated in the same manner. The supernatant was then incubated with 0.1 volume of 3 M sodium acetate (Bioneer, South Korea), 1 µl glycogen (20 mg/ml, Sigma-Aldrich, USA), and 0.7 volume of isopropanol (Sigma-Aldrich, USA) at −20 °C for 2 h, followed by centrifugation at 13,000 × g for 20 min at 4 °C. The pellet was washed with 70% cold ethanol and dried for 20 min at room temperature. Total VLP DNA was eluted in 200 µl RNase-free dH₂O (Bioneer, South Korea) by incubating at 55 °C for 5 min. The VLP DNA was further purified using QIAamp MinElute Virus Spin Kit (Qiagen, Germany). Negative controls were prepared to detect possible contamination from the reagents in the commercial kits, but no DNA was detected. Sequencing libraries were constructed using the Accel-NGS 1S Plus DNA Library kit (Swift Biosciences, USA), according to the manufacturer’s instructions. Shotgun metagenome sequencing was performed on an Illumina NovaSeq 6000 sequencer (2 × 150 bp) (Illumina, USA).

Bacterial DNA extraction and amplicon sequencing

The pellet obtained from the sample homogenate was resuspended in 5 ml SM buffer, and was filtered through 5.0-µm syringe filters (Sartorius, Germany). The filtrate was centrifuged at 8000 × g for 10 min at 4 °C, and bacterial DNA was extracted from the pellet using QIAamp® Fast DNA Stool Mini Kit (Qiagen, Germany) with repeated bead beating at 5.5 m/s for 30 s in a FastPrep®-24 instrument (MP biomedicals, USA). Negative controls were also included to detect possible contamination from the reagents in commercial kits. The V5-V6 regions of bacterial 16S rRNA gene were amplified with chloroplast-excluding primers, combined with Illumina sequencing overhang adapters using AccuPower PCR Premix (Bioneer, South Korea), as described previously⁹. Amplicons were prepared in triplicates and purified using the QIAquick PCR purification kit (Qiagen, USA). The 16S amplicons were sequenced on an Illumina MiSeq sequencer (2 × 300 bp) (Illumina, USA).

Preprocessing of VLP metagenome data

Adapter sequences and low-quality reads were removed from raw reads using Fastp v0.23.1⁸⁴ with default settings. Twenty base pairs were trimmed from the end of R1 reads and the start of R2 reads, according to the manufacturer’s recommendation (Swift Biosciences, USA). Duplicates were removed using Fastp v0.23.1 with the setting “--dup_calc_accuracy 6”. Reads considered being derived from plants were removed by mapping to representative genomes of fresh vegetables (broccoli, GCF_000695525.1; perilla, GCA_004303085.1; spinach, GCF_002007265.1; lettuce, GCF_002870075.1) using Bowtie2 with --local mode. Reads that were taxonomically assigned into the domains Metazoan and Viridiplantae using Kraken2⁸⁵ were further removed. High-quality reads were aligned to the T4 phage genome (HM137666) using Bowtie2 with default settings. The aligned reads and the remaining reads were separated and assembled individually for each sample using metaSPAdes v3.13.1⁸⁶.

Viral operational taxonomic unit

Large contigs (≥2.5 kb) were retained from the assemblies. These contigs were classified as viral, if they met any of the following criteria: (1) classified into categories 1, 2, 4, or 5 by VirSorter v1.0.6⁸⁷; (2) identified as viruses by VirFinder v1.2.1⁸⁸ with a score ≥0.9 and p < 0.05; (3) classified into categories 3 or 6 by VirSorter and identified as viruses by VirFinder with a score ≥0.7 and p < 0.05; (4) classified into categories 3 or 6 by VirSorter and identified as viruses by VIBRANT v1.2.1; (5) identified as viruses by both VirFinder with a score ≥0.7 and p < 0.05, and VIBRANT; or (6) classified as viruses or unclassified with less than 40% open reading frames assigned to Bacteria, Archaea, or Eukaryotes using CAT v4.8.1⁸⁹. VirSorter and VIBRANT were executed in –virome mode. Circular contigs, where the start and end of the contig were identical, were also classified as viral. Exact duplicates were removed using CD-HIT-EST v4.8.1. The completeness of non-redundant viral contigs (complete, high-, medium- and low-quality) were assessed using CheckV v0.7.0. Viral contigs were then clustered into vOTUs using the ANIm of dRep v3.0.1⁹⁰, if they shared ≥95% average nucleotide identity over 85% of the alingment fraction. Representative genomes for the vOTUs were selected based on completeness and centrality. High-quality reads were mapped to the representative genomes using BBMap v38.70 with default parameters, and the genomes with ≥75% breadth of coverage and ≥1× depth of coverage were retained. The number of mapped reads was normalized using the FPKM method, with normalized values rounded down. Alpha-diversity was estimated using the number of vOTUs and Shannon index. Beta-diversity was measured based on Bray-Curtis dissimilarity with log-transformation.

Viral enrichment estimation

Viral enrichment was assessed using ViromeQC with “-w environmental” option. The viral enrichment score was calculated at each stage of sequence processing (raw reads, preprocessed reads, high-quality reads, reads mapped to large contigs, and reads mapped to viral contigs) by counting the number of reads corresponding to large and small subunit ribosomal RNA gene, as well as single-copy universal markers of bacterial housekeeping genes.

Bacterial 16S rRNA gene analysis

16S rRNA gene amplicon sequences were analyzed using QIIME2 v2021.08. Briefly, Illumina adapter sequences were trimmed, and ASVs were generated using DADA2⁹¹. DADA2 corrected low-quality bases and errors, merged paired-end reads, and removed chimeric sequences. Taxonomic identification of the ASVs was performed using Naïve Bayes classifier trained on non-redundant (99% identity) SILVA small subunit rRNA database (version 138). ASVs assigned to eukaryotes, archaea, mitochondria, chloroplast, or unclassified were excluded. Sequencing depth was evenly rarefied to 20,000 sequences, and ASVs with fewer than 10 sequences were further removed. Alpha- and beta-diversity were analyzed in the same manner as the vOTU analysis.

Quantification of total viral and bacterial abundances

The number of total viral genomes was calculated from the vOTU table using FPKM values, with the following equation: Total number of viral genomes = (Number of T4 phage VLPs × ∑ FPKM values of viral contigs)/ ∑ FPKM values of T4 phage contigs. The total number of bacterial genomes was estimated by quantifying 16S rRNA gene copies using quantitative PCR with chloroplast-excluding primers⁹². The 16S rRNA gene sequences of E. coli K12 were used as a standard. The number of 16S rRNA gene copies was then normalized based on the average number of 16S rRNA gene copies per genome in Gammaproteobacteria⁹³ whose ASVs were the most abundant (51.6 ± 28.2% on average) in the bacterial microbiome of fresh vegetables.

Viral clusters and taxonomic identification

The viral genomes from NCBI RefSeq Viral Genome (2022-03-10 release) and IMG/VR (2020-10-12 release) were selected by comparing them with the representative sequences of the vOTUs (BLASTn, E-value < 10⁻⁵). A gene-sharing network was constructed with all vOTU genomes using vConTACT2, with an inflation of 1.5 for protein clusters and 1 for VCs. The abundance of VCs was determined by summing the abundances of all vOTUs within each VC. The genus-level taxonomy of VCs (≥two members) was determined based on the taxonomy of viral genomes from NCBI RefSeq Viral Genome and IMG/VR. Specifically, the protein sequences of the vOTUs were compared with those from NCBI RefSeq Viral Genome and IMG/VR within each VC using BLASTp (E-value < 10⁻⁵, bit-score > 50). The best BLAST results were then used to identify the genus-level taxonomy through the lowest common ancestor algorithm in MEGAN6. vOTUs categorized as “Clustered” were used to determine genus-level taxonomy. vOTUs categorized as “Clustered”, “Clustered/Singleton,” or “Overlap” were considered for family-level taxonomy. The majority rule was employed to determine the family-level taxonomy of vOTUs assigned to the “Overlap” category comprising multiple VCs. The family-level taxonomy of VCs categorized as “Clustered/Singleton” followed the taxonomy of their pre-VCs. The taxonomy of vOTUs were described according to the ICTV taxonomy (2019 release)³³.

Viral lifestyle prediction

The infection lifestyle was predicted using BACPHLIP⁹⁴ and CheckV. Temperate phages were identified by detecting lysogeny-associated genes, such as integrase and recombinase, using BACPHLIP, or by finding integrated viral sequences using CheckV. vOTUs that did not show lysogenic signals were classified as virulent phages.

Prokaryotic host prediction

First, bacterial hosts were predicted using iPHoP v1.3.2 which employs random forest classifier trained on CRISPR spacer identification, k-mer similarity, and homologous sequence match with provided Sept_2021_pub database based on positive predictive value (designated as confidence score ≥ 90). When multiple hosts were predicted, one with the highest score was selected. Subsequently, CRISPR spacer identification and tRNA sequence similarity were employed to identify bacterial hosts for vOTUs when host predictions from iPHoP were not available. CRISPR spacer sequences from Dion et al.⁹⁵ were used to query vOTUs using BLASTn (≥95% identity, ≥95% coverage, and ≤2 mismatches). Additional CRISPR spacer sequences were obtained by extracting them from the bacterial genomes of GTDB (release 214)⁹⁶ and GEM databases⁹⁷ using MinCED v0.4.2⁹⁸ with default parameters, and used in the same manner. tRNA sequences were identified from the genomes of vOTUs using tRNAscan-SE v1.4 with default parameters. The tRNA sequences were compared with bacterial genomes from the GTDB and GEM databases using BLASTn (≥100% identity, ≥100% coverage). Hosts were assigned to vOTUs only if the taxonomy of all hits was consistent. Hosts were not assigned if they were not observed in the 16S rRNA gene data.

Auxiliary metabolic gene annotation

AMGs were identified using DRAM-v and VIBRANT. For DRAM-v annotation, VirSorter2 was run on all vOTUs with “--prep-for-dramv” option to identify viral genes and viral hallmark genes. The results were then used for DRAM-v analysis. Putative AMGs were selected with high auxiliary scores (category 1–3). For VIBRANT annotation, AMGs were identified from vOTUs with a high enrichment of viral genes. Only putative AMGs categorized as metabolic pathway and sulfur relay system in KEGG annotations were considered. Based on the report of Luo et al.²⁰, AMGs related to ribosomal protein, glycosyl transferase, glycoside hydrolase, methyltransferase, peptidase, nucleotide and amino acid metabolism were further excluded, as these genes are essential for viral functioning and are likely to be illegitimate AMGs.

Comparison with public gut virome databases and human gut VLP metagenome datasets

Raw reads of human gut VLP metagenome datasets (13 studies, n = 1306 subjects) were collected from NCBI Sequence Read Archive (Supplementary Data 5). Illumina adapter sequences and low-quality reads were trimmed using Trimmomatic v0.39 (ILLUMINACLIP:TruSeq3-PE-2.fa:2:30:10, LEADING:20, TRAILING:20, SLIDINGWINDOW:4:20, MINLEN:100-240). Exact duplicates were removed using PRINSEQ v0.20.4. Reads mapped to NCBI RefSeq human genome (GCF_000001405) were removed using Bowtie2. High-quality reads were then mapped to the genome sequences of vOTUs and crAssphage (JQ995537) using BBMap v38.70 with settings of 95% identity and “ambiguous=all.” vOTUs with fewer than 10 paired-end reads, or with a depth of coverage of less than 0.5× across all samples were considered potential false positives. The number of mapped reads per vOTU was normalized using the FPKM method, accounting for the read length of each study. Alpha- and beta-diversity were analyzed in the same manner as the vOTU analysis. vOTUs that were significantly enriched or depleted in either healthy or diseased group were identified using a generalized linear-mixed effects model in MaAsLin2⁹⁹. Viral genomes from the Gut Phage Database⁴³, human Gut Virome Database⁴⁴, and Metagenomic Gut Virus⁴⁵ and IMG/VR databases were compared at the genus-level with those of vOTUs using vConTACT2. Species-level comparisons between viral genomes from these databases and vOTUs were determined by calculating average nucleotide values using “anicalc.py” in CheckV.

Associations between dietary patterns and gut-associated vegetable vOTUs

We recruited a total of 53 healthy individuals (Supplementary Data 7), where their dietary intake patterns were evaluated using both short-term and long-term methods (The Institutional Review Board of Chungnam National University, 202012-SB-169-01). We have obtained informed consent from all participants. For short-term assessment, participants reported all foods and beverages consumed over a period of 3 days. Fecal samples were collected on the subsequent day following the completion of the final meal record, and their fecal samples were processed in the same manner as the experimental procedure of the vegetable virome. The meal records were categorized into 16 food groups by modifying the classification system provided by the Rural Development Administration in Korea, and then converted to serving sizes in accordance with the established protocols¹⁰⁰. For long-term assessment, participants completed a food frequency questionnaire comprising 127 food items, designed specifically for Korean adults¹⁰¹. The questionnaire was evaluated according to the frequency and serving size of each item consumed over the past 12 months, subsequently classifying the 127 food items into 20 food groups. The intake amounts of the food groups were transformed into centered log ratios. Correlations between food groups and gut-associated vOTUs were determined using PERMANOVA and db-RDA analysis. The intake values of four short-term food groups (vegetables, fruits, potatoes, and legumes) were further categorized into three groups (high, medium, and low), whose sample number was evenly distributed across the groups. Gut-associated vOTUs that were enriched or depleted in each group were identified using MaAsLin2, with adjustments for fruit, potato, and legume.

Statistics and reproducibility

P-value of <0.05 was considered significant. One-tailed and two-tailed Mann-Whitney U tests, one-way ANOVA with Tukey’s post-hoc test, and Kruskal-Wallis test with Dunn’s post-hoc test were performed using GraphPad Prism v5.01 for Windows (GraphPad Software, USA). PCoA was performed based on Jaccard distance and Bray-Curtis dissimilarity using R package “vegan.” Statistical significance for distance matrices was assessed using the “adonis” function of “vegan” package, with 999 permutations. The best-fit models constrained to host and environmental factors were selected using ordiR2step with forward selection in R package “vegan,” with 999 permutations. The percentages of explained variances for host and environmental factors were calculated using db-RDA of R package “vegan.” vOTUs differentially abundant among groups were determined using MaAsLin2 with a prevalence of ≥10%. No statistical methods were used to predetermine the sample size due to the exploratory nature of the study. No sequencing data obtained from the samples were excluded from the analyses.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The sequencing data for the vegetable virome/amplicon and Korean gut virome generated in this study have been deposited in the EMBL-EBI database under accession codes PRJEB79316 and PRJEB87348, respectively. The sequencing data for the human gut VLP metagenomes used in this study are available in the EMBL-EBI database under accession codes PRJEB7772, PRJEB42612, PRJEB29491, and PRJEB9524, and in the NCBI database under accession codes PRJNA545408, PRJNA504921, PRJNA648797, PRJNA723467, PRJNA722819, PRJNA389927, PRJNA662107, PRJNA891875, and PRJNA622386. The genomes of human gut phages constructed in previous studies were downloaded, as follows: GVD, https://datacommons.cyverse.org/browse/iplant/home/shared/iVirus/Gregory_and_Zablocki_GVD_Jul2020/; MGV, https://portal.nersc.gov/MGV/; GPD, http://ftp.ebi.ac.uk/pub/databases/metagenomics/genome_sets/gut_phage_database/; and IMG/VR3, https://img.jgi.doe.gov/cgi-bin/vr/main.cgi. Source data are provided with this paper.

References

Carter, M. M. et al. Ultra-deep sequencing of Hadza hunter-gatherers recovers vanishing gut microbes. Cell 186, 3111–3124.e13 (2023).
Article PubMed PubMed Central Google Scholar
Moraïs, S. et al. Cryptic diversity of cellulose-degrading gut bacteria in industrialized humans. Science 383, eadj9223 (2024).
Article PubMed PubMed Central Google Scholar
Rothschild, D. et al. Environment dominates over host genetics in shaping human gut microbiota. Nature 555, 210–215 (2018).
Article ADS PubMed CAS Google Scholar
Valles-Colomer, M. et al. The person-to-person transmission landscape of the gut and oral microbiomes. Nature 614, 125–135 (2023).
Article ADS PubMed PubMed Central CAS Google Scholar
Desai, M. S. et al. A dietary fiber-deprived gut microbiota degrades the colonic mucus barrier and enhances pathogen susceptibility. Cell 167, 1339–1353.e21 (2016).
Article PubMed PubMed Central CAS Google Scholar
Sonnenburg, E. D. et al. Diet-induced extinctions in the gut microbiota compound over generations. Nature 529, 212–215 (2016).
Article ADS PubMed PubMed Central CAS Google Scholar
Leff, J. W. & Fierer, N. Bacterial communities associated with the surfaces of fresh fruits and vegetables. PLoS ONE 8, e59310 (2013).
Article ADS PubMed PubMed Central CAS Google Scholar
Kim, M.-S. & Park, E.-J. Composition and variability of core phyllosphere fungal mycobiota on field-grown broccoli. Environ. Microbiome 18, 15 (2023).
Article PubMed PubMed Central Google Scholar
Kim, M.-S., Bae, J.-W. & Park, E.-J. Geographic and host-associated variations in bacterial communities on the floret surfaces of field-grown broccoli. Appl. Environ. Microbiol. 84, e02837–02817 (2018).
Article PubMed PubMed Central Google Scholar
Dees, M. W., Lysøe, E., Nordskog, B. & Brurberg, M. B. Bacterial communities associated with surfaces of leafy greens: shift in composition and decrease in richness over time. Appl. Environ. Microbiol. 81, 1530–1539 (2015).
Article ADS PubMed PubMed Central CAS Google Scholar
Kim, M.-S. & Park, E.-J. Postharvest-induced microbiota remodeling increases fungal diversity in the phyllosphere mycobiota of broccoli florets. Postharvest Biol. Technol. 181, 111693 (2021).
Article CAS Google Scholar
Kim, M. S., Bae, J. W. & Park, E. J. Postharvest processing decreases the richness of bacterial taxa in the phyllosphere of broccoli. J. Appl. Microbiol. 125, 295–305 (2018).
Article PubMed CAS Google Scholar
Berg, G. et al. Plant microbial diversity is suggested as the key to future biocontrol and health trends. FEMS Microbiol. Ecol. 93, fix050 (2017).
Article Google Scholar
Kusstatscher, P. et al. Microbiome approaches provide the key to biologically control postharvest pathogens and storability of fruits and vegetables. FEMS Microbiol. Ecol. 96, fiaa119 (2020).
Article PubMed CAS Google Scholar
Milani, C. et al. Colonization of the human gut by bovine bacteria present in Parmesan cheese. Nat. Commun. 10, 1286 (2019).
Article ADS PubMed PubMed Central Google Scholar
Pasolli, E. et al. Large-scale genome-wide analysis links lactic acid bacteria from food with the gut microbiome. Nat. Commun. 11, 2610 (2020).
Article ADS PubMed PubMed Central CAS Google Scholar
Wicaksono, W. A. et al. The edible plant microbiome: evidence for the occurrence of fruit and vegetable bacteria in the human gut. Gut Microbes 15, 2258565 (2023).
Article PubMed PubMed Central Google Scholar
Cobián Güemes, A. G. et al. Viruses as winners in the game of life. Annu. Rev. Virol. 3, 197–214 (2016).
Article PubMed Google Scholar
Hsu, B. B. et al. Dynamic modulation of the gut microbiota and metabolome by bacteriophages in a mouse model. Cell Host Microbe 25, 803–814.e5 (2019).
Article PubMed PubMed Central CAS Google Scholar
Luo, X.-Q. et al. Viral community-wide auxiliary metabolic genes differ by lifestyles, habitats, and hosts. Microbiome 10, 190 (2022).
Article PubMed PubMed Central Google Scholar
Sinha, A. et al. Transplantation of bacteriophages from ulcerative colitis patients shifts the gut bacteriome and exacerbates the severity of DSS colitis. Microbiome 10, 105 (2022).
Article PubMed PubMed Central CAS Google Scholar
Ritz, N. L. et al. The gut virome is associated with stress-induced changes in behaviour and immune responses in mice. Nat. Microbiol. 9, 359–376 (2024).
Article PubMed PubMed Central Google Scholar
Roux, S. et al. Towards quantitative viromics for both double-stranded and single-stranded DNA viruses. PeerJ 4, e2777 (2016).
Article PubMed PubMed Central Google Scholar
Norman, J. M. et al. Disease-specific alterations in the enteric virome in inflammatory bowel disease. Cell 160, 447–460 (2015).
Article PubMed PubMed Central CAS Google Scholar
Kim, K.-H. & Bae, J.-W. Amplification methods bias metagenomic libraries of uncultured single-stranded and double-stranded DNA viruses. Appl. Environ. Microbiol. 77, 7663–7668 (2011).
Article ADS PubMed PubMed Central CAS Google Scholar
Nayfach, S. et al. CheckV assesses the quality and completeness of metagenome-assembled viral genomes. Nat. Biotechnol. 39, 578–585 (2021).
Article PubMed CAS Google Scholar
Zolfo, M. et al. Detecting contamination in viromes using ViromeQC. Nat. Biotechnol. 37, 1408–1412 (2019).
Article PubMed CAS Google Scholar
Patel, A. et al. Virus and prokaryote enumeration from planktonic aquatic environments by epifluorescence microscopy with SYBR Green I. Nat. Protoc. 2, 269–276 (2007).
Article PubMed CAS Google Scholar
López-García, P. et al. Metagenome-derived virus-microbe ratios across ecosystems. ISME J. 17, 1552–1563 (2023).
Article PubMed PubMed Central Google Scholar
Knowles, B. et al. Lytic to temperate switching of viral communities. Nature 531, 466–470 (2016).
Article ADS PubMed CAS Google Scholar
Lindow, S. E. & Brandl, M. T. Microbiology of the phyllosphere. Appl. Environ. Microbiol. 69, 1875–1883 (2003).
Article ADS PubMed PubMed Central CAS Google Scholar
Roux, S. et al. Minimum information about an uncultivated virus genome (MIUViG). Nat. Biotechnol. 37, 29–37 (2019).
Article PubMed CAS Google Scholar
Adriaenssens, E. M. Phage diversity in the human gut microbiome: a taxonomist’s perspective. mSystems 6, https://doi.org/10.1128/msystems.00799-21 (2021).
Bin Jang, H. et al. Taxonomic assignment of uncultivated prokaryotic virus genomes is enabled by gene-sharing networks. Nat. Biotechnol. 37, 632–639 (2019).
Article Google Scholar
Roux, S. et al. iPHoP: an integrated machine learning framework to maximize host prediction for metagenome-derived viruses of archaea and bacteria. PLoS. Biol. 21, e3002083 (2023).
Article PubMed PubMed Central CAS Google Scholar
Shaffer, M. et al. DRAM for distilling microbial metabolism to automate the curation of microbiome function. Nucleic Acids Res. 48, 8883–8900 (2020).
Article PubMed PubMed Central CAS Google Scholar
Kieft, K., Zhou, Z. & Anantharaman, K. VIBRANT: automated recovery, annotation and curation of microbial viruses, and evaluation of viral community function from genomic sequences. Microbiome 8, 1–23 (2020).
Article Google Scholar
Yuan, Y. et al. Discovery of novel bacterial queuine salvage enzymes and pathways in human pathogens. Proc. Natl. Acad. Sci. USA 116, 19126–19135 (2019).
Article ADS PubMed PubMed Central CAS Google Scholar
Kredich, N. M. Biosynthesis of cysteine. EcoSal Plus 3, https://doi.org/10.1128/ecosalplus.3.6.1.11 (2008).
Zhang, T. et al. RNA viral community in human feces: prevalence of plant pathogenic viruses. PLoS Biol. 4, e3 (2006).
Article PubMed Google Scholar
David, L. A. et al. Diet rapidly and reproducibly alters the human gut microbiome. Nature 505, 559–563 (2014).
Article ADS PubMed CAS Google Scholar
Van Espen, L. et al. A previously undescribed highly prevalent phage identified in a Danish Enteric Virome Catalog. mSystems 6, https://doi.org/10.1128/msystems.00382-21 (2021).
Camarillo-Guerrero, L. F., Almeida, A., Rangel-Pineros, G., Finn, R. D. & Lawley, T. D. Massive expansion of human gut bacteriophage diversity. Cell 184, 1098–1109.e9 (2021).
Article PubMed PubMed Central CAS Google Scholar
Gregory, A. C. et al. The gut virome database reveals age-dependent patterns of virome diversity in the human gut. Cell Host Microbe 28, 724–740.e8 (2020).
Article PubMed PubMed Central CAS Google Scholar
Nayfach, S. et al. Metagenomic compendium of 189,680 DNA viruses from the human gut microbiome. Nat. Microbiol. 6, 960–970 (2021).
Article PubMed PubMed Central CAS Google Scholar
Li, J., Yang, F., Xiao, M. & Li, A. Advances and challenges in cataloging the human gut virome. Cell Host Microbe 30, 908–916 (2022).
Article PubMed CAS Google Scholar
Pinto, Y., Chakraborty, M., Jain, N. & Bhatt, A. S. Phage-inclusive profiling of human gut microbiomes with Phanta. Nat. Biotechnol. 42, 651–662 (2024).
Article PubMed CAS Google Scholar
Wan, Y. et al. Alterations in fecal virome and bacteriome virome interplay in children with autism spectrum disorder. Cell Rep. Med. 5, 101409 (2024).
Article PubMed PubMed Central CAS Google Scholar
Dutilh, B. E. et al. A highly abundant bacteriophage discovered in the unknown sequences of human faecal metagenomes. Nat. Commun. 5, 4498 (2014).
Article ADS PubMed CAS Google Scholar
Shkoporov, A. N. et al. The human gut virome is highly diverse, stable, and individual specific. Cell Host Microbe 26, 527–541.e5 (2019).
Article PubMed CAS Google Scholar
Almeida, A. et al. A unified catalog of 204,938 reference genomes from the human gut microbiome. Nat. Biotechnol. 39, 105–114 (2021).
Article PubMed CAS Google Scholar
Nayfach, S., Shi, Z. J., Seshadri, R., Pollard, K. S. & Kyrpides, N. C. New insights from uncultivated genomes of the global human gut microbiome. Nature 568, 505–510 (2019).
Article ADS PubMed PubMed Central CAS Google Scholar
Hannigan, G. D. et al. Diagnostic potential and interactive dynamics of the colorectal cancer virome. mBio 9, https://doi.org/10.1128/mbio.02248-18 (2018).
Zuo, T. et al. Gut mucosal virome alterations in ulcerative colitis. Gut 68, 1169–1179 (2019).
Article PubMed CAS Google Scholar
Lang, S. et al. Intestinal virome signature associated with severity of nonalcoholic fatty liver disease. Gastroenterology 159, 1839–1852 (2020).
Article PubMed CAS Google Scholar
Yang, K. et al. Alterations in the gut virome in obesity and type 2 diabetes mellitus. Gastroenterology 161, 1257–1269.e13 (2021).
Article PubMed CAS Google Scholar
Fan, G. et al. Alterations in the gut virome are associated with type 2 diabetes and diabetic nephropathy. Gut Microbes 15, 2226925 (2023).
Article PubMed PubMed Central Google Scholar
Nayak, B., Liu, R. H. & Tang, J. Effect of processing on phenolic antioxidants of fruits, vegetables, and grains—a review. Crit. Rev. Food Sci. Nutr. 55, 887–918 (2015).
Article PubMed CAS Google Scholar
Carmody, R. N., Varady, K. & Turnbaugh, P. J. Digesting the complex metabolic effects of diet on the host and microbiome. Cell 187, 3857–3876 (2024).
Article PubMed CAS Google Scholar
Reyes, A. et al. Viruses in the faecal microbiota of monozygotic twins and their mothers. Nature 466, 334–338 (2010).
Article ADS PubMed PubMed Central CAS Google Scholar
Manrique, P. et al. Healthy human gut phageome. Proc. Natl. Acad. Sci. USA 113, 10400–10405 (2016).
Article ADS MathSciNet PubMed PubMed Central CAS Google Scholar
Kim, M.-S. & Bae, J.-W. Lysogeny is prevalent and widely distributed in the murine gut microbiota. ISME J. 12, 1127–1141 (2018).
Article PubMed PubMed Central CAS Google Scholar
Liang, G. et al. The stepwise assembly of the neonatal virome is modulated by breastfeeding. Nature 581, 470–474 (2020).
Article ADS PubMed PubMed Central CAS Google Scholar
Pei, Z. et al. A universe of human gut-derived bacterial prophages: unveiling the hidden viral players in intestinal microecology. Gut Microbes 16, 2309684 (2024).
Article PubMed PubMed Central Google Scholar
Kumbhari, A. et al. Discovery of disease-adapted bacterial lineages in inflammatory bowel diseases. Cell Host Microbe 32, 1147–1162.e12 (2024).
Article PubMed CAS Google Scholar
Carmody, R. N. et al. Cooking shapes the structure and function of the gut microbiome. Nat. Microbiol. 4, 2052–2063 (2019).
Article PubMed PubMed Central Google Scholar
Delaroque, C. et al. Diet standardization reduces intra-individual microbiome variation. Gut Microbes 14, 2149047 (2022).
Article PubMed PubMed Central Google Scholar
Motta, J.-P., Wallace, J. L., Buret, A. G., Deraison, C. & Vergnolle, N. Gastrointestinal biofilms in health and disease. Nat. Rev. Gastroenterol. Hepatol. 18, 314–334 (2021).
Article PubMed Google Scholar
Rossi, E. et al. “It’s a gut feeling”—Escherichia coli biofilm formation in the gastrointestinal tract environment. Crit. Rev. Microbiol. 44, 1–30 (2018).
Article PubMed CAS Google Scholar
Chang, Y.-C., Lo, H.-H., Hsieh, H.-Y. & Chang, S.-M. Identification, epidemiological relatedness, and biofilm formation of clinical Chryseobacterium indologenes isolates from central Taiwan. J. Microbiol. Immunol. Infect. 48, 559–564 (2015).
Article PubMed CAS Google Scholar
Luis, A. S., Yates, E. A. & Cartmell, A. Functions and specificity of bacterial carbohydrate sulfatases targeting host glycans. Essays Biochem. 67, 429–442 (2023).
Article PubMed Google Scholar
Anderson, C. L., Sullivan, M. B. & Fernando, S. C. Dietary energy drives the dynamic response of bovine rumen viral communities. Microbiome 5, 1–19 (2017).
Article Google Scholar
de Jonge, P. A. et al. Gut virome profiling identifies a widespread bacteriophage family associated with metabolic syndrome. Nat. Commun. 13, 3594 (2022).
Article ADS PubMed PubMed Central Google Scholar
Clooney, A. G. et al. Whole-virome analysis sheds light on viral dark matter in inflammatory bowel disease. Cell Host Microbe 26, 764–778. e765 (2019).
Article PubMed CAS Google Scholar
Liao, C. et al. Oral bacteria relative abundance in faeces increases due to gut microbiota depletion and is linked with patient outcomes. Nat. Microbiol. 9, 1555–1565 (2024).
Article PubMed PubMed Central CAS Google Scholar
von Hertzen, L. Plant microbiota: implications for human health. Br. J. Nutr. 114, 1531–1532 (2015).
Article Google Scholar
Willner, D. et al. Metagenomic detection of phage-encoded platelet-binding factors in the human oral cavity. Proc. Natl. Acad. Sci. USA 108, 4547–4553 (2011).
Article ADS PubMed CAS Google Scholar
Boling, L. et al. Dietary prophage inducers and antimicrobials: toward landscaping the human gut microbiome. Gut Microbes 11, 721–734 (2020).
Article PubMed PubMed Central Google Scholar
Zmora, N. et al. Personalized gut mucosal colonization resistance to empiric probiotics is associated with unique host and microbiome features. Cell 174, 1388–1405.e21 (2018).
Article PubMed CAS Google Scholar
Han, G., Luong, H. & Vaishnava, S. Low abundance members of the gut microbiome exhibit high immunogenicity. Gut Microbes 14, 2104086 (2022).
Article PubMed PubMed Central Google Scholar
Wortelboer, K. et al. Phage-microbe dynamics after sterile faecal filtrate transplantation in individuals with metabolic syndrome: a double-blind, randomised, placebo-controlled clinical trial assessing efficacy and safety. Nat. Commun. 14, 5600 (2023).
Article ADS PubMed PubMed Central CAS Google Scholar
Tian, X. et al. Gut virome-wide association analysis identifies cross-population viral signatures for inflammatory bowel disease. Microbiome 12, 130 (2024).
Article PubMed PubMed Central CAS Google Scholar
Bonilla, N. et al. Phage on tap–a quick and efficient protocol for the preparation of bacteriophage laboratory stocks. PeerJ 4, e2261 (2016).
Article PubMed PubMed Central Google Scholar
Chen, S., Zhou, Y., Chen, Y. & Gu, J. fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics 34, i884–i890 (2018).
Article PubMed PubMed Central Google Scholar
Wood, D. E., Lu, J. & Langmead, B. Improved metagenomic analysis with Kraken 2. Genome Biol. 20, 1–13 (2019).
Article Google Scholar
Nurk, S., Meleshko, D., Korobeynikov, A. & Pevzner, P. A. metaSPAdes: a new versatile metagenomic assembler. Genome Res. 27, 824–834 (2017).
Article PubMed PubMed Central CAS Google Scholar
Roux, S., Enault, F., Hurwitz, B. L. & Sullivan, M. B. VirSorter: mining viral signal from microbial genomic data. PeerJ 3, e985 (2015).
Article PubMed PubMed Central Google Scholar
Ren, J., Ahlgren, N. A., Lu, Y. Y., Fuhrman, J. A. & Sun, F. VirFinder: a novel k-mer based tool for identifying viral sequences from assembled metagenomic data. Microbiome 5, 1–20 (2017).
Article Google Scholar
Von Meijenfeldt, F. B., Arkhipova, K., Cambuy, D. D., Coutinho, F. H. & Dutilh, B. E. Robust taxonomic classification of uncharted microbial sequences and bins with CAT and BAT. Genome Biol. 20, 1–14 (2019).
Google Scholar
Olm, M. R., Brown, C. T., Brooks, B. & Banfield, J. F. dRep: a tool for fast and accurate genomic comparisons that enables improved genome recovery from metagenomes through de-replication. ISME J. 11, 2864–2868 (2017).
Article PubMed PubMed Central CAS Google Scholar
Callahan, B. J. et al. DADA2: high-resolution sample inference from Illumina amplicon data. Nat. Methods 13, 581–583 (2016).
Article PubMed PubMed Central CAS Google Scholar
Rastogi, G., Tech, J. J., Coaker, G. L. & Leveau, J. H. A PCR-based toolbox for the culture-independent quantification of total bacterial abundances in plant environments. J. Microbiol. Methods 83, 127–132 (2010).
Article PubMed CAS Google Scholar
Větrovský, T. & Baldrian, P. The variability of the 16S rRNA gene in bacterial genomes and its consequences for bacterial community analyses. PLoS ONE 8, e57923 (2013).
Article ADS PubMed PubMed Central Google Scholar
Hockenberry, A. J. & Wilke, C. O. BACPHLIP: predicting bacteriophage lifestyle from conserved protein domains. PeerJ 9, e11396 (2021).
Article PubMed PubMed Central Google Scholar
Dion, M. B. et al. Streamlining CRISPR spacer-based bacterial host predictions to decipher the viral dark matter. Nucleic Acids Res. 49, 3127–3138 (2021).
Article PubMed PubMed Central CAS Google Scholar
Parks, D. H. et al. GTDB: an ongoing census of bacterial and archaeal diversity through a phylogenetically consistent, rank normalized and complete genome-based taxonomy. Nucleic Acids Res. 50, D785–D794 (2022).
Article PubMed CAS Google Scholar
Nayfach, S. et al. A genomic catalog of Earth’s microbiomes. Nat. Biotechnol. 39, 499–509 (2021).
Article PubMed CAS Google Scholar
Bland, C. et al. CRISPR recognition tool (CRT): a tool for automatic detection of clustered regularly interspaced palindromic repeats. BMC Bioinform. 8, 1–8 (2007).
Article Google Scholar
Mallick, H. et al. Multivariable association discovery in population-scale meta-omics studies. PLoS Comput. Biol. 17, e1009442 (2021).
Article PubMed PubMed Central CAS Google Scholar
Yoon, S.-I. et al. Nutrient inadequacy in Korean young adults with depression: a case control study. Nutrients 15, 2195 (2023).
Article PubMed PubMed Central CAS Google Scholar
Yun, S. H., Shim, J.-S., Kweon, S. & Oh, K. Development of a food frequency questionnaire for the Korea National Health and Nutrition Examination Survey: data from the fourth Korea National Health and Nutrition Examination Survey (KNHANES IV). Korean J. Food Nutr. 46, 186–196 (2013).
Article Google Scholar

Download references

Acknowledgements

This research was supported by the National Research Foundation of Korea (NRF) grants funded by the Ministry of Education (2018R1D1A3B07050366 to E.J.P.) and the Ministry of Science and ICT (MSIT) (No. 2019R1C1C1009664 to M.S.K., No. 2021R1F1A1064222 to E.J.P., and No. 2020R1A5A8017671 to M.S.K.).

Author information

Authors and Affiliations

Department of Microbiology and Molecular Biology, Chungnam National University, Daejeon, Republic of Korea
Ji-Woo Park, Yeo-Eun Yun, Su-A In & Min-Soo Kim
Department of Food and Nutrition, Chungnam National University, Daejeon, Republic of Korea
Jin Ah Cho & Su-In Yoon
Department of Food Bioengineering, Jeju National University, Jeju, Republic of Korea
Eun-Jin Park

Authors

Ji-Woo Park
View author publications
Search author on:PubMed Google Scholar
Yeo-Eun Yun
View author publications
Search author on:PubMed Google Scholar
Jin Ah Cho
View author publications
Search author on:PubMed Google Scholar
Su-In Yoon
View author publications
Search author on:PubMed Google Scholar
Su-A In
View author publications
Search author on:PubMed Google Scholar
Eun-Jin Park
View author publications
Search author on:PubMed Google Scholar
Min-Soo Kim
View author publications
Search author on:PubMed Google Scholar

Contributions

J.W.P., E.J.P., and M.S.K. conceived and designed the experiments; J.W.P., Y.E.Y., and S.A.I. performed the experiments; J.W.P., Y.E.Y., and M.S.K. analyzed the data; J.A.C, S.I.Y., E.J.P., and M.S.K. contributed materials/analysis tools; J.W.P., E.J.P., and M.S.K. wrote the manuscript.

Corresponding authors

Correspondence to Eun-Jin Park or Min-Soo Kim.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Description of Additional Supplementary Files

Supplementary Data 1

Supplementary Data 2

Supplementary Data 3

Supplementary Data 4

Supplementary Data 5

Supplementary Data 6

Supplementary Data 7

Reporting Summary

Transparent Peer Review file

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Park, JW., Yun, YE., Cho, J.A. et al. Characterization of the phyllosphere virome of fresh vegetables and potential transfer to the human gut. Nat Commun 16, 3427 (2025). https://doi.org/10.1038/s41467-025-58856-6

Download citation

Received: 06 October 2024
Accepted: 31 March 2025
Published: 10 April 2025
Version of record: 10 April 2025
DOI: https://doi.org/10.1038/s41467-025-58856-6