2. Materials and methods

2.1. Strains and culture conditions

All Halophytophthora isolates/strains investigated in this study were collected in December 2015 from seven different marine and brackish water sites along the Algarve coast of southern Portugal using an in situ baiting method (Jung et al. 2017) in salt marshes, tidal pools, channels, lagoons, and estuaries that remained submerged even at low tide (Maia et al. 2022). Supplementary Table S1 provides detailed information and species identification for all Halophytophthora isolates included in the experiments. All strains were cultured on salty V8-agar medium (sV8A; 16 g agar, 3 g CaCO3, 100 ml Campbell’s V8 juice, 450 ml distilled water, 450 ml seawater (Maia et al. 2022); or liquid V8-medium containing seawater adjusted to 50% concentration (sV8).

2.2. RNA extraction

Total RNA was purified from ~100 mg of 7–10-days old mycelium using RNAzol® RT Column Kit (Chomczynski et al. 2010) and treated with TURBO DNA-free™ Kit (Ambion). RNA was quantified in Qubit® 2.0 Fluorometer (Invitrogen), and quality was tested by Tape Station 4200 (Agilent). One RNA pool was prepared with eight isolates derived of the original ex-type species of Halophytophthora described in Maia et al. (2022).

2.3. RNA library preparation and total stranded RNA sequencing

Around 1 μg of total RNA, eluted in RNase-free water, was sent to IABio in Olomouc, Czech Republic, for RNA library construction and deep sequencing. Ribosomal RNA (rRNA) was depleted using the NEBNext rRNA Depletion Kit (Human/Mouse/Rat) and prepared with the NEBNext Ultra II Directional RNA Library Prep Kit for Illumina, along with NEBNext Multiplex Oligos for Illumina (Unique Dual Index Primer Pairs). Library quality control was evaluated with the Agilent Bioanalyzer 2100 High Sensitivity DNA Kit. The KAPA Library Quantification Kit for the Illumina platform facilitated absolute quantitative PCR (qPCR)-based quantification of the Illumina libraries, flanked by the P5 and P7 flow cell oligo sequences. The libraries underwent paired-end (PE) (2 × 150 nt) sequencing on a NovaSeq6000 (DS-150) from Illumina, San Diego, CA, USA, using the NovaSeq S4 v1.5 reagent kit. An ‘in-lane’ PhiX control spike was included in each lane of the flow cell.

2.4. Virus identification bioinformatics pipeline

2.4.1. Preprocessing of RNA-seq data

Raw data were processed by BaseSpace cloud interface (Illumina) in default settings. The processing was carried out on the local server of the University of South Bohemia in České Budějovice, Czech Republic. The basecalling, adapter clipping, and quality filtering were carried out using Bcl2fastq v2.20.0.422 Conversion Software (Illumina). For the initial quality assessment, we used FastQC v0.11.9 (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/). The adapter sequences, provided by the sequencing firm, were verified using BBTools v37.87 (https://jgi.doe.gov/data-and-tools/software-tools/bbtools/) (Bushnell et al. 2017). Following processing with Cutadapt v4.7 (https://cutadapt.readthedocs.io/en/stable/) (Martin 2011) involved trimming of N bases, adapter sequences, and low-quality ends (<30). Reads falling short of 50 bp were discarded from subsequent analyses. Post-trimming quality was again re-assessed with FastQC. The remaining rRNA reads were removed using SortMeRNA v4.3.6 (https://github.com/sortmerna/sortmerna/) (Kopylova et al. 2012). SILVA SSU and LSU databases filtered to include only Oomycota rRNA sequences were used as references (https://www.arb-silva.de/).

2.4.2. Alignment/read mapping to host genome

For mapping reads to the host genome, the STAR v2.7.9a (Dobin 2014) program was employed with default settings. Two genomes, H. batemanensis (GCA_023338225.1) and

H. polymorphica (GCA_023338205.1), were used as references. After alignment to the reference sequence, all mapped reads were discarded and only unmapped reads were utilized further.

2.4.3. De novo assembly of unmapped reads for detection of novel viruses

The unmapped reads were used for de novo assembly. To assemble these reads into contigs, the SPAdes program v3.15.5 (https://github.com/ablab/spades/) (Bankevich et al. 2012) was used with default settings for metagenomics. For further analysis, assembled contigs shorter than 500 bp were discarded.

2.4.4. Search for similarity in virus databases

Final contigs were cross-referenced against several databases using the Basic Local Alignment Search Tool (BLAST) v2.10.0 (https://blast.ncbi.nlm.nih.gov/doc/blast-help/downloadblastdata.html/ (Camacho et al. 2009). As a reference, we used viral reference sequences from NCBI. BLASTn was applied for the nucleotide database accessible at https://www.ncbi.nlm.nih.gov/labs/virus/vssi/#/virus?SeqType_s=Nucleotide. Similarly, BLASTx was employed for the protein database available at https://www.ncbi.nlm.nih.gov/labs/virus/vssi/#/virus?SeqType_s=Protein and also to search the UniProt database for a specified virus taxon at https://www.uniprot.org/uniprot/?query=taxonomy:10239. Contigs with an e-value higher than 1e-3 were discarded. Subsequently, the remaining contigs were processed in a similar manner using the entire range of databases. For BLASTn, the NCBI nucleotide (nt) database from NCBI, available at https://ftp.ncbi.nlm.nih.gov/blast/db/nt*, was used. For BLASTx, both the NCBI nr database, accessible at https://ftp.ncbi.nlm.nih.gov/blast/db/nr*, and the UniProt Knowledgebase (UniProtKB) available at https://www.uniprot.org/uniprotkb/, were employed. The BLAST search was restricted to the best hit per contig. The final list of candidate viral contigs was compiled by scanning the BLAST results for the search string ‘vir’.

To identify homologous replicases and hypothetical proteins that might not be annotated in Blastx, a BLASTn search against the NCBI nt database using the full-length HthCRESSV1 genomic sequence as a query. Hits with high nucleotide similarity (e-value <1e−10) were retrieved and manually inspected. Top matches corresponded to mitochondrial genome assemblies from diverse oomycete species. To confirm the genomic context and sequence identity, the HthCRESSV1-Rep and HP sequences were mapped to these mitochondrial genome assemblies using Geneious Prime (v2023.2.2), employing medium sensitivity and default parameters.

2.4.5. Coverage depth

The number of unique sequencing reads that align to each reference de novo assembled viral contig was calculated with the following formula: (Total reads mapped to the final identified virus * average read length)/virus genome or contig length) (Sims et al. 2014). Coverage plots were visualized with Integrated Genome Viewer (IGV) tool and Geneious Prime® 2025.0.2. (Figure S1).

2.5. DNA extraction and PCR reactions

Purified DNA used for Southern hybridization was extracted using the modified cetyltrimethylammonium bromide (CTAB)-based method by Fukuhara et al. (1993). Specifically, ~0.1 g of dried mycelia harvested after 7 days of culture in sV8 liquid medium was used as starting material, and the extraction was performed on a small scale. DNA extraction from mycelia after single lesion isolation derived from zoospores was performed as described by Maia et al. (2022). DNA was stored at −80°C for long-term preservation. PCR reactions for complete sequence determination and detection of HtCRESSV1 were performed using the obtained DNA as template with primers listed in Table S2 and GoTaq (Promega, Wisconsin, USA).

2.6. Phylogenetic analyses

Maximum likelihood (ML) phylogenetic trees were constructed using RAxML-HPC v.8 on XSEDE via the CIPRES Science Gateway (Stamatakis 2006). The analysis employed a rapid bootstrapping algorithm with the recommended parameters, and the GAMMA model was used to avoid extensive optimization of the best-scoring ML tree at the end of the run. The Jones–Taylor–Thornton (JTT) model was applied as the amino acid (aa) substitution model for protein sequences.

The resulting trees were visualized using iTOL v6 (Letunic and Bork 2021) (https://itol.embl.de).

2.7. Stem loop prediction

DNA secondary structure prediction and free energy calculations were carried out using two web-based computational tools. The mFold web server (hosted by UNAFold) was used to predict and visualize DNA secondary structures and determine associated thermodynamic parameters, including minimum free energy, under defined temperature and ionic conditions (https://www.unafold.org/mfold/applications/Structure-display-and-free-energy-determination.php). Additionally, the ViennaRNA Web Services platform (http://rna.tbi.univie.ac.at/) was also employed for further structure prediction and analysis. Although primarily designed for RNA, the ViennaRNA tools were adapted for DNA by specifying the appropriate nucleic acid type and parameters where applicable.

2.8. Analysis of the ORF2 protein sequence and structure

Secondary structure and putative transmembrane domains were predicted using the PSIPRED and MEMSAT-SVM algorithms, respectively, available through the PSIPRED Workbench (Buchan et al. 2024). Coiled-coil regions were predicted using PCOILS available through the MPI Bioinformatics Toolkit (Zimmermann et al. 2018). Structural modelling of ORF2 protein sequence was carried out using AlphaFold3 (Abramson et al. 2024). The models were built for different oligomeric states from monomer to homo-decamer. The quality of the models was evaluated by comparing their predicted template modelling (pTM) and the interface predicted template modelling (ipTM) scores. The resulting models were visualized using ChimeraX v1.11.1 (Meng et al. 2023).

2.9. Virus-like particle purification and transmission electron microscopy observation

Approximately 40 g of BD651 mycelium cultured in liquid s. V8 medium was collected using Miracloth, washed with distilled water, and incubated in Buffer A (0.1 M phosphate buffer, 0.2 M KCl) for 10 min. The mycelium was collected, homogenized in liquid nitrogen, and disrupted using a French press. Triton X-100 (1% final concentration) was added to the cell lysate, followed by stirring at 4°C for 30 min. The lysate was centrifuged at 1000 g for 10 min to remove cellular debris. The supernatant was collected, mixed with an equal volume of chloroform, gently agitated for 30 min, and centrifuged at 8000 g for 15 min. The aqueous phase was collected and centrifuged at 20 000 g for 30 min to pellet heavy membrane fractions. The resulting supernatant was ultracentrifuged at 100 000 g for 1 h using a Hitachi CP80WX P45AT rotor to pellet virus-like particles (VLPs; P100 000). The pellet was resuspended in Buffer A and subjected to continuous sucrose density gradient centrifugation (10%–40% w/v) at 80 000 g for 3 h at 4°C using a Hitachi CP80WX P28S swing rotor. Following ultracentrifugation, the gradient was fractionated from the top in 2 ml aliquots using a Hitachi DGF-U gradient fractionator, yielding 15 fractions. Total nucleic acids were extracted from each fraction, and HthCRESSV1-enriched fractions (fractions 4 and 5) were identified by HthCRESSV1-specific PCR. These fractions were pooled and concentrated by ultracentrifugation at 100 000 g for 1 h using a Hitachi CP80WX P45AT rotor. The pellet was resuspended in phosphate buffer, negatively stained with 2% uranyl acetate, and visualized by transmission electron microscopy (TEM) using a JEM-1400 Plus microscope (JEOL, Tokyo, Japan).

2.10. Southern blot analysis

Approximately 5 μg of purified DNA extracted from strain BD651 were subjected to electrophoresis on a 0.8% agarose gel, followed by depurination treatment by immersing the gel in 0.25 M HCl and shaking for 5 min. After washing the gel with distilled water, DNA was denatured by shaking in denaturing solution (0.5 M NaOH, 1.5 M NaCl) for 15 min. This operation was performed twice. The denatured gel was then neutralized by shaking in neutralizing solution (0.5 M Tris–HCl, pH 7.5, 1.5 M NaCl) for 15 min, repeated twice, washed with distilled water, and equilibrated by incubation in 20× SSC for 10 min. Subsequently, the DNA was transferred to a positively charged nylon membrane (Roche, Basel, Switzerland) by capillary blotting. To detect the HtCRESSV1 genome, a DIG-labelled DNA probe was prepared using the HtCRESSV1-HP probe (nt 114–837) according to the protocol provided with the PCR DIG Labeling Mix (Roche, Basel, Switzerland). Hybridization with the DNA probe was performed at 50°C for 16 h. Subsequent operations until detection were performed according to the method of Sakuta et al. (2023).

2.11. Mitochondrial purification and nucleic acid extraction

Approximately 20 g of mycelia were homogenized in 200 ml of isotonic solution (0.25 M sucrose, 100 mM Tris–HCl, pH 7.4, 10 mM EGTA). The homogenate was transferred to a 50 ml centrifuge tube and centrifuged at 1000 × g for 10 min to precipitate cell debris and nuclei (P1000). The supernatant was transferred to a Hizacks tube and centrifuged at 12 000 × g for 15 min to obtain a crude mitochondrial fraction enriched in mitochondria (P12 000). This supernatant was used as a membrane fraction and centrifuged at 100 000 × g for 90 min using a P45AT rotor in a Hitachi CP80WX (Hitachi Koki, Japan), with the resulting pellet designated as the membrane fraction (P100 000) and the supernatant as the cytosolic fraction (S100 000). P12 000 was purified using two rounds of sucrose density gradient centrifugation to obtain a more intact mitochondrial fraction. First, it was layered on a discontinuous sucrose solution prepared at concentrations of 15%–23%–32%–60% and centrifuged at 82 200 × g for 3 h using a P28S swing rotor in a Hitachi CP80WX (Hitachi Koki, Japan). After centrifugation, the purified mitochondrial concentrated fraction observed at the 32%–60% interface was carefully collected, diluted with isotonic solution, centrifuged at 10 000 × g for 20 min, and the resulting pellet was resuspended in isotonic solution to obtain a semi-purified mitochondrial fraction. The second purification was performed by layering the semi-purified mitochondrial fraction on a sucrose solution prepared at 32%–60% concentration and ultracentrifuging under the same conditions as the first round. After centrifugation, 2 ml fractions were carefully collected from the top to obtain 15 fractions (F1–F15). Total DNA was extracted from the 15 fractions containing purified mitochondrial fractions using a modified method of Okada et al. (2015). Specifically, samples were mixed with 0.5 ml of 2× Saline-Tris-EDTA (STE) buffer (200 mM NaCl, 20 mM Tris–HCl pH 8.0, 2 mM EDTA pH 8.0) containing 1% sodium dodecyl sulfate (SDS) and 0.5 ml of 1:1 phenol/chloroform. The mixture was vortexed and centrifuged at 15 000 g for 5 min. The resulting supernatant was used as a total nucleic acid sample, and after ethanol precipitation, it was suspended in TE (1M Tris-HCl, 0.5M EDTA, pH 8.0) buffer containing RNase A and incubated at 37°C for 30 min to obtain a total DNA sample.

2.12. COX activity measurement

To measure the activity of cytochrome c oxidase, a mitochondrial marker enzyme, the Cytochrome c Oxidase Assay Kit (Sigma) was used according to the manufacturer’s instructions. Enzyme activity was measured by monitoring the absorbance at 550 nm for 1 min, with the decrease in absorbance defined as ΔA and substituted into the following equation to measure enzyme activity: Units/ml·g = ΔA/min × 1.1/(0.1 × 21.84 × X) (X: weight of source mycelia in grams).

2.13. Zoospore induction and single lesion isolation, mycelial comparison, and growth tests

Zoospore induction of H. thermoambigua was performed using modified protocols from Jesus et al. (2016) and Jung et al. (1999). Single lesion formation was achieved by floating young cork oak (Quercus suber) leaves on the water surface in nonsterile 50% seawater after zoospore induction and incubating at 20°C under natural light for 24 h. The single lesions were excised with a razor blade and placed on PARPNH agar medium (V8 juice agar (V8A) supplemented with 10 μg/ml pimaricin, 200 μg/ml ampicillin, 10 μg/ml rifampicin, 25 μg/ml pentachloronitrobenzene (PCNB), 50 μg/ml nystatin, and 50 μg/ml hymexazol) to obtain single zoospore-derived isolates. The relationship between temperature and growth was examined by preculturing target strains on 90 mm V8A medium for 1 week to standardize the growth stage. Subsequently, three plates per isolate were incubated at 4°C, 8°C, 15°C, 20°C, 25°C, 28°C, 30°C, and 33°C in the dark. The test was conducted for 5 days or until the colonies nearly reached the edge of the Petri dish, and the average growth rate (cm/day) was calculated.

2.14. Attempted HthCRESSV1 elimination

A total of 20 single-zoospore strains derived from the isolate BD651 were prepared, following the procedure described by Uchida et al. (2021) and Sakuta et al. (2025). The regeneration medium was supplemented with ribavirin (300 μg ml−1 final concentration) and cycloheximide (5 μg ml−1 final concentration), and 20 isolates were subsequently recovered from the regenerated mycelia. Total nucleic acids were then extracted from each culture grown on sV8 medium using the phenol–chloroform–isoamyl alcohol (PCI) method. The presence of the virus was finally verified by PCR employing a primer set specific to HtCRESSV1 (Table S2).

2.15. Statistical analysis

Colony diameter measurement data were subjected to analysis of variance. Dunnett’s test following one-way analysis of variance was applied for comparison of means between virus-infected and noninfected strain groups. Statistical significance was set at P < .05 (*P < .05, **P < .01).