Crop management shapes the diversity and activity of DNA and RNA viruses in the rhizosphere

Abstract

Background The rhizosphere is a hotspot for microbial activity and contributes to ecosystem services including plant health and biogeochemical cycling. The activity of microbial viruses, and their influence on plant-microbe interactions in the rhizosphere, remains undetermined. Given the impact of viruses on the ecology and evolution of their host communities, determining how soil viruses influence microbiome dynamics is crucial to build a holistic understanding of rhizosphere functions. Results Here, we aimed to investigate the influence of crop management on the composition and activity of bulk soil, rhizosphere soil, and root viral communities. We combined viromics, metagenomics, and metatranscriptomics on soil samples collected from a 3-year crop rotation field trial of oilseed rape (Brassica napus L.). By recovering 1059 dsDNA viral populations and 16,541 ssRNA bacteriophage populations, we expanded the number of underexplored Leviviricetes genomes by > 5 times. Through detection of viral activity in metatranscriptomes, we uncovered evidence of “Kill-the-Winner” dynamics, implicating soil bacteriophages in driving bacterial community succession. Moreover, we found the activity of viruses increased with proximity to crop roots, and identified that soil viruses may influence plant-microbe interactions through the reprogramming of bacterial host metabolism. We have provided the first evidence of crop rotation-driven impacts on soil microbial communities extending to viruses. To this aim, we present the novel principal of “viral priming,” which describes how the consecutive growth of the same crop species primes viral activity in the rhizosphere through local adaptation. Conclusions Overall, we reveal unprecedented spatial and temporal diversity in viral community composition and activity across root, rhizosphere soil, and bulk soil compartments. Our work demonstrates that the roles of soil viruses need greater consideration to exploit the rhizosphere microbiome for food security, food safety, and environmental sustainability.

Crop rotation significantly influences the composition of soil, rhizosphere, and root microbiota in canola (Brassica napus L.)

“…While other members of the genus Olpidium , in particular Olpidium virulentis , are known vectors for plant viruses [ 51 ], widespread reclassification within this genus means the potential of O. brassicae to harbour and transmit viruses that may impact canola yield requires further investigation. In this context, it is germane to note that rhizosphere, root, and bulk soil feature very high spatial and temporal diversity in virome composition in B. napus , with continuous canola providing a viral “priming” function that results in an increase in viral load and species diversity [ 52 ]. It is possible, though unproven, that such priming could be mediated by the selection of various strains of O. brassicae as a result of high frequency canola cropping strategies.…”

Protein embeddings improve phage-host interaction prediction

“…To identify the putative gene products, we configured Prokka [50] to refer to PHROG [52], a database of viral protein family clusters generated by employing hidden Markov model profile-profile comparisons for remote homology detection. PHROG [52] has also been used in previous studies that require the functional annotation of phage protein sequences [53, 54]. The annotated RBPs were selected following the same scheme described in the previous paragraph.…”

Mesophilic and thermophilic viruses are associated with nutrient cycling during hyperthermophilic composting

“…The non-redundant functional proteins in viral contigs were annotated (Table S8 ) using VIBRANT based on Pfam, dbCAN, KEGG and eggNOG databases 5.0 with default parameters [ 4 , 61 ]. The phage lifestyle was predicted using three tools including VIBRANT [ 56 ], PhaTYP [ 62 ] and manually curated BLAST [ 25 ] based on previously described methods [ 25 , 63 ]. Briefly, VIBRANT [ 56 ] and manual BLAST [ 25 ] were used to infer temperate lifestyle by identifying viral contigs that contained proteins associated with lysogeny (transposase, integrase, excisionase, resolvase, and recombinase) [ 74 ].…”

“…The phage lifestyle was predicted using three tools including VIBRANT [ 56 ], PhaTYP [ 62 ] and manually curated BLAST [ 25 ] based on previously described methods [ 25 , 63 ]. Briefly, VIBRANT [ 56 ] and manual BLAST [ 25 ] were used to infer temperate lifestyle by identifying viral contigs that contained proteins associated with lysogeny (transposase, integrase, excisionase, resolvase, and recombinase) [ 74 ]. Due to incompleteness of several viral contigs, a machine learning method called PhaTYP [ 62 ] was also used to predict the phage lifestyle.…”

“…For example, it has been shown that viral AMGs can potentially help host bacteria to acquire nutrients and degrade toxic compounds, having positive effect on their survival [ 14 ]. Given that viruses require host cell machinery for the transcription of their genes, viral activity based on transcriptomics can be used to indicate successful infection of hosts [ 25 ]. This approach has recently been used to demonstrate that sub-seafloor sediments show activity by both lytic and lysogenic viruses [ 26 ], that the expression of AMGs are possibly involved in modulating host methane and sulfur metabolism upon infection in ocean ecosystem [ 22 , 27 ] and that giant viruses are active in coastal marine system [ 28 ].…”