Genomics discovery of giant fungal viruses from subsurface oceanic crustal fluids

The study investigates the diversity and ecology of eukaryotic Nucleocytoviricota (giant dsDNA viruses) in the subsurface oceanic crust, an environment known to harbor diverse bacteria, archaea, viruses, and fungi. Prior work at the Juan de Fuca Ridge flank detected primarily archaeal and bacterial viruses in crustal fluids, with only minor signals attributed to NCLDV-like genes that were interpreted as archaeal. Nucleocytoviricota infect a broad range of eukaryotes but remain underexplored in this biome. The research question centers on whether eukaryotic NCLDVs are present and active in oceanic crustal fluids, their taxonomic placement, genomic features, and potential hosts. The purpose is to detect, assemble, and characterize NCLDV genomes from single virions and infer host associations, with implications for microbial ecology and biogeochemical cycling in the marine deep biosphere.

Background studies describe the presence of diverse microbial life (bacteria, archaea, fungi) and viruses in oceanic sediments and crustal fluids at the Juan de Fuca Ridge flank. Earlier metagenomic surveys largely recovered archaeal and bacterial viruses, with limited and ambiguous evidence for NCLDVs. Nucleocytoviricota are a monophyletic group of large dsDNA viruses infecting many eukaryotes, with wide diversity revealed by metagenomics and single-virus genomics, but with limited understanding of their roles in subsurface crustal ecosystems. Advances in flow cytometry sorting and single-virus sequencing have enhanced recovery of NCLDV genomes from environmental samples, motivating the present targeted approach.

• Sample site and collection: Crustal fluid sampled July 2011 at IODP borehole U1362B (47.754997 N, 127.4753 W), ~2650 m water depth; fluids from 359 m below seafloor (basaltic basement). One-milliliter aliquot preserved with 5% glycerol at −80 °C.
• Flow cytometry and sorting: Thawed sample diluted 1:100 in 0.2 µm-filtered 1× TE, stained with SYBR Green I. Virus-like particles (VLPs) and cells visualized and sorted (BD Influx). Individual VLPs sorted into half of a 384-well plate; bacterial/archaeal cells into the other half. Controls included non-sort wells.
• Whole genome amplification and screening: Each well lysed with 0.4 M KOH for 10 min; DNA amplified via multiple displacement amplification (MDA). Plate screened by PCR for bacterial/archaeal 16S rRNA to exclude cellular contamination; 72 putative vSAGs (MDA-positive, 16S-negative) identified; 8 randomly selected for sequencing (vSAG1v1-8), with technical replicate libraries for some.
• Library preparation and sequencing: Illumina 300 bp paired-end libraries prepared and sequenced on MiSeq.
• Read processing and assembly: Reads quality-trimmed with Trimmomatic (slidingwindow:8:15, MINLEN:36) and low-complexity reads removed (BBTools). Assemblies generated with SPAdes v3.15.0 using careful and metaviral modes; contigs filtered by coverage and repaired as needed.
• Annotation and contamination assessment: Initial annotations via BLAST against NCBI-nr (e-value ≤ 1e-10), with additional checks using RefSeq and RVDB. Two vSAGs (vSAG1-J6R and vSAG1-J6F) identified as eukaryotic NCLDVs. Scaffolds 100% identical to Marine Alphaproteobacterial sequences (non-viral) were flagged as contaminants and removed. Gene organization visualized with Easyfig. Putative bacterial, archaeal, and eukaryotic gene annotations cataloged.
• Phylogenetic analyses: Identification of conserved NCLDV marker genes (including DNA polymerase B, D5-like primase-helicase, packaging ATPase, LTF3-like, superfamily I/II helicases). Proteins aligned with MAFFT; low-information positions removed; concatenated alignments analyzed with IQ-TREE and FastTree using best-fit models. ANI comparisons computed (FastANI). eIF4E and tRNATyr genes analyzed phylogenetically; tRNA genes predicted with tRNAscan-SE/ARAGORN, intron positions assessed. Presence/absence of nucleocytoplasmic virus core gene groups (NVCGs/NCLVOGs) mapped.
• Eukaryotic community assessment: Taxonomic assignment of 18S rRNA, 28S rRNA, and ITS sequences from U1362B (and U1362A) metagenomes via BLAST against SILVA to determine eukaryotic taxa present and their likely hosts.

• Two related draft NCLDV genomes (vSAG1.JdFr and vSAG8.JdFr) were recovered from a 1 mL crustal fluid sample; they share 99.8% ANI with 70% alignment fraction, indicating closely related strains.
• These genomes were not found in contemporaneous metagenomes from U1362A/B, suggesting rarity or undersampling in bulk metagenomics; nevertheless, IMG/M indicated ~3.31% of sequences in crustal fluids correspond to NCLDVs.
• Phylogenomics placed vSAG1.JdFr and vSAG8.JdFr within Mesomimiviridae, related to Phycodnaviruses (PhyS) and Chrysochromulina parva virus B02 (CpV B02), with ~65–67% ANI to these references.
• Core gene content: At least 11 and 13 ancestral NCLDV core gene groups (NVCGs/NCLVOGs) detected in vSAG1.JdFr and vSAG8.JdFr, respectively; presence/absence patterns matched Mesomimiviridae profiles.
• Gene repertoire indicates horizontal gene transfer: Both genomes contain clusters of genes annotated as bacterial, archaeal, and eukaryotic, organized in genomic islands alongside NCLDV genes; archaeal annotations predominantly affiliated with marine lineages, including Euryarchaeota detected in crustal fluids.
• Eukaryotic gene content points to fungal association: Both genomes encode posttranslational modification enzymes (e.g., ubiquitin-conjugating enzyme, mitochondrial iron-sulfur transporter Atm1), peroxisome biogenesis factor 10 (variant 2), and a regulatory protein MGI1; phylogenies support fungal origins via HGT.
• Translation-related genes: Both genomes encode a tyrosine tRNA (tRNATyr) with a short intron (~20 bp) at the canonical 37/38 position, a hallmark of eukaryotic/archaeal tRNA introns; both also encode eukaryotic translation initiation factor 4E (eIF4E), clustering with Mesomimiviridae and other giant viruses.
• Environmental context: Ascomycota fungi were the only confirmed eukaryotes in U1362B crustal fluids based on near-complete 18S rRNA, partial 28S rRNA, and ITS sequences; other eukaryotic fragments matched plants/algae/animals but are unlikely residents of the crustal environment.
• Collectively, evidence supports Ascomycota fungi as the likely hosts for these NCLDVs, representing a putative fungi–Nucleocytoviricota association.

Targeted single-virion genomics revealed two Mesomimiviridae NCLDVs in subsurface oceanic crustal fluids that were missed by bulk metagenomics, likely due to rarity or sequencing depth limitations. The genomes carry bacterial and archaeal gene islands indicative of gene flow within the host's ecological niche and suggest interactions among fungi, bacteria, and archaea in the basaltic basement. Fungal-origin eukaryotic genes, the presence of translation-related components (tRNATyr with a short canonical intron and eIF4E), and the co-occurrence of Ascomycota as the only confirmed eukaryotes in the sampled crustal fluids together support fungi—most likely Ascomycota—as the putative hosts. This extends the known host range and ecology of NCLDVs into the oceanic crustal biosphere and implies that viral infection of fungi could influence productivity and biogeochemical processes (e.g., carbon and nitrogen cycling, hydrogen production supporting chemolithotrophs) in this anoxic environment. The finding aligns with prior hints of NCLDV-fungal interactions and suggests that ancestral fungal lineages may harbor NCLDV infections.

The study reports the discovery and genomic characterization of two closely related Mesomimiviridae NCLDVs from subsurface oceanic crustal fluids at Juan de Fuca Ridge. Genomic features, including fungal-derived genes via HGT, tRNATyr with a short canonical intron, and eIF4E, together with environmental community data, implicate Ascomycota fungi as putative hosts—representing a first putative fungi–Nucleocytoviricota association. These results highlight active host–virus dynamics in the oceanic crust and suggest that NCLDV infections of fungi may be ecologically significant for deep biosphere biogeochemistry. Future work should (i) expand targeted single-virion and metagenomic sampling across crustal sites and times, (ii) obtain more complete viral genomes and assess infection dynamics, (iii) directly link viruses to hosts (e.g., via viral tagging, FISH, CRISPR-spacer or integrated elements), and (iv) evaluate ecological impacts of fungal NCLDV infections on carbon, nitrogen, and hydrogen cycling.

• Potential database misannotations may affect taxonomic assignment of bacterial/archaeal/eukaryotic genes in NCLDV genomes.
• Evidence for host identity is circumstantial (co-occurrence, HGT signatures) without direct infection assays or physical associations.
• Flow cytometry/SYBR Green I approach is biased toward dsDNA viruses; small particles, RNA viruses, and ssDNA mycoviruses likely underdetected or missed.
• Single-virion MDA can introduce amplification biases and chimeras; genomes are draft and incomplete; gene copy numbers and genome organization may be imperfectly resolved.
• Putative contaminant scaffolds were detected and removed; the contamination source was not identified.
• Bulk metagenomes lacked these NCLDVs, possibly due to low sequencing depth or rarity, limiting cross-validation.
• Publication text contains minor inconsistencies in sample identifiers (e.g., vSAG numbering in data availability), reflecting potential reporting errors.