A giant virus awakens polinton-like virophages in the green alga Tetraselmis , revealing an inducible antiviral defense system

Giant double-stranded DNA viruses infecting microalgae often interact with polinton-like viruses (PLVs), yet their dynamics within natural hosts remain poorly understood. Here, we report the isolation of a new T etraselmis striata giant virus (Allomimiviridae) from Mediterranean coastal waters, which we designate Oceanusvirus lionense (T etV-2). From the same sample, we co-isolated a PLV, T sa-S2b, that replicates concomitantly with T etV-2. Remarkably, T etV-2 infection also reactivates at least two additional PLVs integrated into the host genome. We show that these PLVs exert a virophage-like effect, reducing the fitness of the giant virus. Comparative genomics reveals that these virophagic PLVs are related to the autonomous lytic PLV T sV-N1, sharing structural but not replication-associated genes. This newly described pathosystem provides a unique model to investigate the evolution of PLVs and the parasitic mechanisms underlying virophage–giant virus interactions.

The green algal genus T etraselmis ( Chlorodendrophyceae, Chlorophyta) comprises more than 30 species widespread in coastal and estuarine ecosystems worldwide, where they contribute to primary production and occasionally form dense blooms [1– 4]. Several species are of biotechnological importance as aquaculture feed and for biofuel production [5,6]. Genomic and transcriptomic resources are now available for multiple T etraselmis lineages [7–12], providing opportunities to investigate their interactions with viruses. T etraselmis species also serve as hosts for a diversity of double-stranded DNA viruses. The first are giant Oceanusviruses (family Allomimiviridae, order Imitervirales), represented by T etraselmis virus 1 (Oceanusvirus kanoehense), which was isolated off the Hawaiian coast using an unidentified local T etraselmis strain as bait [13,14]. T etV-1 possesses a 226–257 nm icosahedral capsid and a 668-kb circular genome encoding 653 predicted proteins and 10 tRNAs, making it one of the largest known algal viruses. Its genome encodes atypical metabolic enzymes, including those involved in mannitol metabolism, saccharide degradation, and fermentation, and its replication cycle has been described in detail [15]. In contrast, T etraselmis striata virus N1 (T sV-N1), isolated from a Norwegian fjord, represents a distinct group of viruses known as polinton-like viruses (PLVs) [16]. PLVs are small dsDNA viruses with genome sizes of 20–31 kb, encoding capsid proteins related to those of virophages and polintons [17–21]. The T sV- N1 genome, although only partially sequenced (31 kb), encodes 33 predicted proteins and forms ~60 nm icosahedral particles. A closely related isolate from the Black Sea, T. viridis virus S1 (TvV-S1), shares >99% nucleotide identity with T sV-N1 [22]. Both T sV- N1 and TvV-S1 displays a lytic lifestyle, completely clearing host cultures within 3–5 days. Comparative genomics revealed that Chlorodendrophyceae genomes are rich in viral insertions, including large integrated fragments related to both Oceanusvirus and PLVs [23]. Recent evidence suggests that some PLVs can adopt a virophage-like lifestyle. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint The Gezel-14T virus, isolated from Phaeocystis globosa, depends on co-infection with a giant virus for replication, reduces giant virus yield, and integrates into the algal genome [24,25]. These properties mirror those of bona fide virophages (family Lavidaviridae), which parasitize the replication factories of giant viruses [21,26–28]. Whether PLVs integrated in T etraselmis genomes can similarly be reactivated by giant virus infection remains unknown. Here, we report the isolation from Mediterranean coastal waters and complete genome sequencing of a new Oceanusvirus, TetV-2, infecting T. striata. We propose that T etV-2 defines a new species, Oceanusvirus lionense. Importantly, we show that T etV-2 infection triggers the replication of both a co-isolated PLV and a subset of the endogenous PLV elements integrated in the T. striata genome, all closely related to T sV-N1. These T sV-like genomes are assembled into virus-like particles, providing evidence that integrated PLVs in T etraselmis can be reactivated by giant virus infection. Moreover, the T sV-like elements exert a negative impact on T etV-2 replication, indicating a virophage-like lifestyle. T ogether, our findings expand the known repertoire of virophage-like interactions in marine microalgae and underscore the complexity of virus–virus–host dynamics in coastal ecosystems.

Virus isolation and culture A culture of T etraselmis striata was inoculated with 0.45 µm–filtered environmental water collected in May 2022 from the Lion lagoon, southern France. Within the first 24 hours post-infection (h.p.i.), cells exhibited a marked loss of motility accompanied by deflagellation. By 72 h.p.i., the culture had largely cleared, with a pronounced decline in viable cell numbers observed microscopically. T o confirm the stability of the lytic activity and fulfill Koch’s postulates, several successive rounds of infection were carried out by inoculating fresh T. striata cultures with 0.45 µm–filtered lysates. Sequencing of the TetV-2 genome Diagnostic short-read sequencing of the initial 0.45 µm–filtered lysate yielded 20 contigs encoding proteins predominantly most similar to homologs from Oceanusvirus kaneohense (T etV-1). These contigs spanned a total of 632 kb and displayed a relatively uniform sequencing depth of 792× ± 90. The proteins encoded by these contigs matched distinct, non-overlapping regions of the T etV-1 genome, indicating that they likely represent fragments of a single virus genome closely related to T etV-1. However, attempts to reconstruct a complete genome from the short-read dataset, including multiple contig extension trials, were unsuccessful. T o overcome this limitation, we isolated a single viral clone from the lysate, designated T etV-2, through several rounds of serial dilution to extinction, and sequenced its genome using a long-read approach. Assembly produced a single circular contig of 660,680 bp, encoding 640 predicted proteins and 4 tRNAs. A large fraction of these proteins (n = 389; 61%) shared significant similarity with T etV-1 homologs, of which 348 had their best BLAST match in T etV-1 ( Fig. 1A ). Reciprocal best-hit analysis identified 343 orthologous pairs. The four T etV-2 tRNAs corresponded to a subset of the T etV-1 tRNA repertoire (10 genes) and, as in T etV-1, were clustered within a single genomic locus. T etV-2 lacked several genes present in T etV-1, including two fermentation-related genes encoding pyruvate formate-lyase (T etV_428) and its activating enzyme preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint (T etV_456) [13], as well as a photolyase (T etV_298) and a deoxyribodipyrimidine photolyase-related protein (T etV_522), both likely involved in UV-induced DNA repair. Notably, T etV-2 also lacked thymidylate synthase (T etV_240) and thymidylate kinase (T etV_225), which catalyze successive steps in the conversion of dUMP to dTDP . In contrast, 54 T etV-2 proteins had no detectable similarity to T etV-1 but did display significant matches to bacterial, eukaryotic, or viral proteins. Among these were enzymes involved in lipid metabolism, including a very long-chain fatty acid elongase (T etV2_00470) and a lysophospholipid acyltransferase (T etV2_00549). We also identified two genes encoding putative mitochondrial proteins: an “altered inheritance of mitochondria” protein 24 (AIM24; T etV2_00382) and a calcium-binding mitochondrial carrier protein (T etV2_00259). A substantial fraction of T etV-2 proteins (n = 264; 41%) had no detectable homologs in the NR database; these were globally shorter than conserved proteins (average length 200 vs. 377 amino acids, respectively). On average, orthologous proteins shared 49% sequence identity at the amino acid level (Fig. 1B). Figure 1 : Comparative genomics of TetV. (A) T axonomic origin of the best matches in the GenBank NR database to T etV-2 proteins.(B) Sequence identity distribution among reciprocal best matches between T etV-1 and T etV-2 proteins.(C) Maximum likelihood phylogenetic reconstruction (IQ-TREE) based on a concatenated alignment of the A18 helicase, packaging ATPase, and DNA polymerase. The tree was rooted using chlorovirus and mamiellovirus homologs as outgroup sequences (not shown). Bootstrap support values <90% are indicated.(D) Gene colinearity between T etV-1 and T etV-2 genomes. Genes are represented as colored boxes: pink, conserved genes on the same strand; green, conserved genes on opposite strands; purple, genes with significant NR matches but lacking reciprocal best hits between the two viruses; gray, hypothetical genes without significant matches. T riangles shows gene doublets encoding a bacterial transposase- related protein and an uncharacterized “transposase-associated” protein. Phylogenetic placement of T etV-2 within the Allomimiviridae [29] was supported by analyses based on concatenated alignments of three hallmark Megaviricetes proteins: the A18 helicase, packaging ATPase, and DNA polymerase ( Fig. 1C). Despite infecting distinct T etraselmis hosts, T etV-1 and T etV-2 were more closely related to one another TetV-2 TetV-1 TetV-1 (348) Other dsDNA viruses (20) Eukaryotes (34) Bacteria (24) No Hit (214) Faunusvirus sp. Rheavirus sinusmexicani (CroV) Megavirus chiliensis Tupanvirus altamarinense Terrestrivirus sp. Yasminevirus saudimassiliense Tethysvirus hollandense (PgV) Organic Lake phycodnavirus 1 Kratosvirus quantuckense (AaV) Mimiviridae sp. ChoanoV1 Heliosvirus raunefjordenense (PoV) Dishui Lake large algae virus 1 Oceanusvirus kaneohense (TetV-1) Biavirus raunefjordenense (PkV) 72 39 0.50 Tethysvirus ontarioense (CpV) Oceanusvirus lionense (TetV-2) A B C D 10 20 30 40 50 60 70 80 90 1000 10 30 20 Protein identity (%) Frequency (%) TetV-1 vs. TetV-2 Schizo. Allo. Meso. Mimiviridae TetV Other dsDNA viruses Eukaryotes Bacteria No Hit preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint than to any endogenous T etV-like elements integrated within T etraselmis genomes, including those of T. striata , the host of T etV-2 ( Fig. S1 ). This indicates that the diversity of T etraselmis-infecting viruses extends beyond that represented by T etV-1 and T etV-2. Base composition differed substantially between the two viruses: T etV-1 and T etV-2 had GC contents of 41.2% and 57.8%, respectively. Even higher GC levels were observed among endogenous T etV-like elements in T. striata, T. suecica, and T. chui (ranging from 58.8% to 73.9%). These compositional differences are unlikely to reflect adaptive convergence toward host GC content, since host genomes themselves vary only modestly (58% in T. striata vs. 53% in T. suecica and T. chui). Regions of gene colinearity were apparent between T etV-1 and T etV-2 but were frequently interrupted by large-scale genomic rearrangements ( Fig. 1D). The T etV-2 genome contained eight nearly identical copies (97.5–100% nucleotide identity) of a gene doublet encoding a bacterial transposase-related protein and an uncharacterized “transposase-associated” protein. These duplications likely contributed to the failure of short-read assembly. By contrast, T etV-1 contained only a single copy of this doublet. Phylogenetic analyses indicated that the duplications occurred after divergence of the T etV-1 and T etV-2 lineages (Fig. S2). Several T etV-2 doublets were positioned at the boundaries of colinear regions, suggesting a role in mediating homologous recombination and genome rearrangements. Given their close relationship yet distinct host specificities, we propose that T etV-2 be assigned to the genus Oceanusvirus as the first representative of a new species, O. lionense, named after the “salin du Lion” site where it was first isolated. Reactivation and genomic features of endogenous Tsv-like viruses Analysis of the initial short-read assembly revealed five contigs (1,775–17,559 bp) matching endogenous T sv elements (ETEs) previously identified in the T. striata LANL1001 genome, with >99% nucleotide identity across their entire length ( Fig. S3) [23]. Short-read mapping against the LANL1001 reference uncovered ten additional regions with significant coverage, all but one overlapping known ETEs, including four fully spanned between terminal inverted repeats (TIRs; C2208, C1731, C0566, C2184). T wo of these (C2208, C2184) were also supported by long reads from the T etV-2 lysate, indicating that their corresponding viral genomes were present in an independent infection experiment. No other LANL1001 regions were supported by long-read data. Although the BG host strain used here differs from LANL1001, both belong to the same T. striata species group and share near-identical marker genes (100% identity over a 780 bp RBCL fragment; 99.8% identity over a 1,542 bp 18S fragment; Fig. S4 and Supplementary Data). Thus, the observed read recruitment cannot be attributed to host DNA release during lysis, which would yield random mapping patterns. Instead, the data indicate that the lysates contained additional T sv-like viruses closely related to T. striata ETEs. T o reconstruct these genomes, we performed ETE-guided local assemblies for the four fully covered elements. Each yielded a single contig flanked by TIRs, corresponding to full-length T sv-like genomes designated T sv-S2a, T sv-S2a.bis, T sv-S2b, and T sv-S3b (hereafter S2a, S2a.bis, S2b, and S3b; Fig. 2). preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint PCR assays with subtype-specific primers confirmed the presence of S2a and S3b sequences both in uninfected T etraselmis BG host DNA and as encapsidated DNA in filtered lysates, consistent with the reactivation of endogenous elements. In contrast, S2b DNA was detected exclusively in lysates, suggesting that it was co-isolated with T etV-2 from the environmental sample rather than derived from the host genome. This interpretation was further supported by serial passages of purified T etV-2, which consistently yielded lysates containing S2a and S3b but never S2b, demonstrating that S2b is not generated through endogenous element reactivation. S2a.bis could not be assessed, as its sequence lacks unique regions suitable for primer design that would allow discrimination from other T sV variants. A B T etraselmis elements Tsv-S1 mCP mCP ATPase POLD4 VLTF3 ClyA-like 10 Kb Tsv-S2b Tsv-S2a Tsv-S2a.bis Tsv-S3b PLV-RED1 PLV-MED1 PLV-INO1 PLV-SPO1 PLV-SPO2 C3275 C0574 C1502 C0382 C3091 C3079 C1186 C0293 C1731 C2601 C2532 C3146 C2097 Tsv-S1 C1307 C0694 C0826 C0660 C3212 C3296 C0668 C2184 Tsv-S3b C0552 C0674 C1893 C1127 C1893 C3053 C3128 C2288 C2624 C1585 Tsv-S2a.bis Tsv-S2a C0566 C2208 C1731 Tsv-S2b C0942 0.50 ANI: 95.7% ANI: 62.2% SLATTSAP pDNAP TFP MCP D5HP TFP MCP D5HP TFP MCP D5HP MCPTFP S1H Smc FA58CTFPMCP YR YR YR 19,063 bp 19,170 bp 19,124 bp 18,038 bp ANI: 99.3% Figure 2. Tetraselmis striata Tsv genomes (A) Genome maps of T etraselmis striata viruses. Genes with annotated functions are shown as colored rectangles; hypothetical genes are shown in gray. Genes with significant protein-level similarity (BLASTP) are connected by lines, and regions with significant nucleotide similarity (BLASTN) are highlighted in green. Pink lines indicate genes conserved across all five viruses. T erminal inverted repeats are marked in red at the genome extremities. Average nucleotide identities (ANI) between genomes were calculated as the mean across BLASTN high-scoring pairs (HSPs). (B) Maximum-likelihood phylogenetic tree of the MCPs. Branch support values correspond to approximate likelihood-ratio test (aLRT) local supports (only values <80 are shown). PLV sequences used as outgroups (RED1, MED1, INO1, SPO1, and SPO2) were obtained from Yutin et al. (2015). T sv-like elements detected by lysate sequencing are highlighted in red. Abbreviations: POLD4, DNA polymerase delta subunit 4; VLTF3, very late transcription factor 3; S1H, superfamily 1 helicase fused to an inactivated pDNAP domain; mCP, minor capsid protein; ATPase, DNA packaging ATPase; Smc, structural maintenance of chromosomes domain; Clya-like, ClyA-like α-pore-forming toxin domain; MCP, major capsid protein; TFP, tail fiber protein; FA58C, coagulation factor 5/8 C-terminal (discoidin) domain; D5HP, helicase–primase D5; SAP, SAP DNA-binding domain protein; AOx, PLN02976 amine oxidase domain; pDNAP, protein-primed family B DNA polymerase; SLATT_4, SMODS and SLOG-associating 2TM effector domain family 4; YR, tyrosine recombinase. Insufficient sequencing depth prevented assembly of the remaining ETEs, but mapping patterns suggested additional low-abundance, likely truncated, T sv-like fragments. The four genomes (18,038–19,170 bp; Table S1) were substantially smaller than the free- living T sV-S1 genome (26,407 bp [16]). TIRs ranged from 583 bp (S3b) to 1,724 bp preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint (S2b) and were absent from T sV-S1. The three S2x genomes (i.e., S2a, S2a.bis and S2b) terminated with monotonous “AT” dinucleotide repeats ( Fig. S5), also present in their ETE guides, whereas S3b and its guide C2184 lacked such repeats. Each contig shared very high nucleotide identity with its ETE guide (ANI = 99.1–100%; Table S2). Notably, S2a.bis was slightly more similar to S2a (ANI = 99.3%) than to its guide C0566 (ANI = 99.1%), suggesting it does not derive from reactivation of a C0566 ortholog in BG. The S2b contig also differed from its guide, which carried a 5,441 bp tandem duplication absent from the viral genome ( Fig. S6 ). By contrast, S3b was more divergent (ANI = 62.2–66.1% with the S2x group), with similarity limited to the distal two-thirds of the genome. No significant nucleotide similarity was detected between T sV-S1 and any contig by BLASTN, although BLASTP readily identified homologous proteins. Phylogenetic analysis of the major capsid protein (MCP) confirmed these relationships (Fig. 2B): the four novel T sv-like viruses clustered more closely with each other than with T sV-S1, forming two clades—S2x and S3b. Short-read-covered LANL1001 ETEs (highlighted in red in Fig. 2B) also grouped into these two clades, indicating that ETE reactivation was not random but restricted to subsets of related elements, possibly responding to regulatory cues triggered by T etV-2 replication. We cannot exclude, however, that fewer T svs were reactivated and that lower/partial coverage of other ETEs reflects spurious read alignments due to sequencing errors. Gene content analysis showed that S2a, S2a.bis, and S3b each encoded 16 predicted proteins, whereas S2b encoded 17. All four shared five conserved genes with T sV-S1: the major and minor capsid proteins, a DNA-packaging ATPase, a tail fiber protein (TFB), and a coagulation factor 5/8 C-terminal domain (FA58C)-like protein likely involved in carbohydrate binding [30]. These conserved genes were clustered on one side of the genome, with clade-specific genes located on the opposite side. Unique regions were enriched in replication functions. T sV-S1 encoded a superfamily 1 helicase fused to an inactive protein-primed DNA polymerase domain (S1H) [31]; S3b encoded a protein-primed family B DNA polymerase (pDNAP) [32]; and the S2x group encoded a D5 helicase–primase (D5HP) adjacent to a tyrosine recombinase, potentially mediating proviral integration and excision [33]. Beyond the core capsid and replication modules, the novel T sVs shared five additional genes absent from T sV-S1, including a SAP DNA-binding protein, a SLATT-domain protein related to predicted membrane-perforating toxins [34], and three uncharacterized proteins. Ultrastructural analysis Transmission electron microscopy (TEM) provided direct evidence of viral infection within T etraselmis cells and revealed the presence of two distinct types of virus-like particles (Fig. S7). Negative staining of cell lysates showed large icosahedral particles measuring 214–239 nm in diameter, consistent in size and morphology with T etraselmis virus 1 (T etV-1; 226–257 nm) and therefore identified as T etV-2 virions. A second population of smaller icosahedral particles, 66–75 nm in diameter, corresponded to the size range previously reported for T etraselmis virus N1 (T sV-N1; 49–73 nm). The coexistence of these two particle classes indicates concurrent production of a giant virus (T etV-2) and smaller T sV-like viruses. T o characterize the replication dynamics of T etV-2, thin-section TEM was performed on T. striata cells at 24, 48, and 72 hours post-inoculation (hpi). At 24 hpi, infected cells displayed extensive cytoplasmic reorganization, with large regions occupied by empty preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint and filled T etV-2 capsids arising from membranous precursors (Fig. S7 A1-C2). These features closely resembled those described for T etraselmis sp. infected with T etV-1 [15], including putative cytoplasmic viral factories with spatially organized capsid maturation, immature particles located centrally and mature virions peripherally. A minority population of smaller icosahedral particles with electron-dense cores was also observed, likely corresponding to T sV-like viruses replicating in parallel with T etV-2. By 48 hpi, the number of T etV-2 particles had markedly declined, while T sV-like particles became predominant, frequently accumulating at the cell periphery near chloroplast membranes rather than in the central cytoplasm where T etV-2 factories had formed (Fig. S7 D1-F2). This shift in viral abundance suggests a temporal succession in which T sV replication becomes predominant after the peak of T etV-2 particle production. At 72 h.p.i., T etV-2 particles were rarely detected, and the cytoplasm was almost entirely filled with T sV-like particles ( Fig. S7 G1-I2 ). These smaller particles often formed crystalline arrays, a hallmark of advanced viral replication and accumulation. Collectively, these ultrastructural observations demonstrate that the polinton-like virus (PLV) genomes identified by metagenomic sequencing correspond to actively replicating viruses that co-occur with, but are distinct from, the giant T etV-2. T o verify encapsidation of the T sV-like viruses, we applied a size-fractionation and DNase-protection assay previously used to demonstrate the virion nature of Gezel-14T [24]. Fresh lysates were sequentially filtered through 0.45-µm and 0.1-µm membranes to remove intact cells and giant viruses, respectively. Aliquots of the raw lysate, clarified (0.45-µm) filtrate, and 0.1-µm filtrate were either left untreated or boiled to release encapsidated nucleic acids, followed by DNase treatment to degrade non- encapsidated DNA. PCR amplification of specific marker genes revealed T etV-2 DNA only in the non-boiled raw and clarified fractions, whereas T sV-like markers were detected in all non-boiled fractions. These results indicate that T etV-2 and T sV-like viruses are independently encapsidated, forming large (>0.1 µm) and small (<0.1 µm) viral particles, respectively. TsVs exhibit a virophage-like lifestyle T o evaluate whether T sVs act as virophages, we investigated their impact on the replication dynamics of the giant virus T etV-2. Host cultures were coinfected with a fixed concentration of T etV-2 (3.10 6 ml -1) and varying ratios of T sV to T etV-2. T wo distinct T sV cocktails derived from independent lysates were tested: one dominated by S2b (i.e., 99.4% S2b, 0.6% S3b; 0,0 % S2a but still detectable by qPCR [est. 8.2 copies per ml]) and the other dominated by S2a (99.4% S2a, 0.6% S3b; S2b undetectable). Cultures were incubated for 13 days, after which both host cells and viral particles were enumerated. Control infections with T sVs alone yielded no detectable T sV progeny (Fig. S8) and did not increase host mortality (Fig. 3A). Instead, T sV abundance declined by more than 98% between inoculation and the end of the experiment, likely reflecting adsorption to host cell surfaces and virion degradation. This confirmed that T sVs cannot replicate autonomously in this host and are not directly harmful to the host. By contrast, coinfections with T etV-2 revealed a strong inhibitory effect of both T sV cocktails on giant virus production (Fig. 3B). Increasing the T sV/T etV-2 ratio in the inoculum caused a log-linear, dose-dependent reduction in T etV-2 yields. These findings indicate that preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint T sVs require T etV-2 for propagation and, in turn, suppress T etV-2 replication, consistent with a virophage-like lifestyle. Nevertheless we did not observe a measurable reduction in host mortality in cultures co-infected with T etV-2 and T sVs; by day 13, all such cultures had visibly clarified. Figure 3. Virophage-associated life traits of TsV. (A) Abundance of T etV-2 particles at 13 days post infection (dpi) as a function of the initial T sV:T etV-2 inoculation ratio. Cultures were co-infected with T etV-2 ( 3 × 10⁶ particles mL⁻¹ ) and a T sV mixture dominated by strain T sV-S2a (red) or T sV-S2b (blue). The T etV-2–only control is shown in gray. Each point represents an independent biological replicate (n = 3), and lines depict the corresponding regression fits. (B) Concentration of viable T etraselmis cells at 13 dpi. Gray bars indicate control cultures: uninfected, infected with the T sV-S2b–dominant mix only (10 × 10⁶ particles mL⁻¹ total T sV), or infected with T etV-2 alone. Red and blue bars correspond to co- infections with T etV-2 and the T sV-S2a– or T sV-S2b–dominant mixes, respectively. Bars represent mean values (n = 3), and error bars denote standard deviations. Replication cycle and latency phase We investigated the replication dynamics of T etV-2 and reactivation of T sv through a triplicate infection experiment, using cultures inoculated with purified T etV-2 particles. Samples were collected over an 8-day period post-inoculation. Host and viral abundances were monitored by flow cytometry, while extracellular encapsidated T sv genome copies were quantified by qPCR. The infection dynamics revealed that the latency phase of T etV-2 last between 21 and 24 hours (Fig. 4). After this period, T etV-2 abundance increased rapidly until ~3 days post-infection (d.p.i.), after which it stabilized and remained constant until the end of the experiment. Detection of the two endogenous polinton-like viruses (T sv-S2a and preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint T sv-S3b) was delayed relative to T etV-2: T sv-S2a was first detected at 24 hours post- infection (h.p.i.), while T sv-S3b appeared only at 2 d.p.i. Unlike T etV-2, both PLVs continued to accumulate throughout the experiment without reaching a plateau. By the final sampling point, T sv-S2a abundance exceeded that of T etV-2 by ~9-fold and that of T sv-S3b by ~452-fold. These quantitative patterns are consistent with TEM observations. At 24 h.p.i., the cytoplasm of infected cells contained predominantly T etV-2 particles, with only a few T sv particles visible, reflecting the initial burst of giant virus replication. By 72 h.p.i., however, giant viral particles were largely absent, and cells were densely filled with small T sv particles. This transition in particle composition mirrors the dynamics observed in abundance profiles: the stabilization of T etV-2 levels after 3 d.p.i. corresponds to the progressive and sustained amplification of the virophages, which continued to accumulate until the end of the experiment. Days post infection 0 1 72 3 4 5 6 8 1M 2.5M 2.0M 1.5M 1.0M 0.5M 10K 100 TetV-2 Tsv-S2a Tsv-S3b T. striata NI T. striata INF 0 1 72 3 4 5 6 8 Virus abundance log(ml-1) Cell abundance (ml-1) Figure 4. Infection dynamics of Tetraselmis by TetV-2. (A) Replication dynamics of T etV-2, T sV-S2a, and T sV-S3b following infection at a ratio of 30 T etV-2 particles per host cell. T etV-2 particle abundance was quantified by flow cytometry, while encapsidated DNA copy numbers of T sV-S2a and T sV-S3b were determined by qPCR. (B) Host cell survival over time, assessed by flow cytometry in infected and uninfected control cultures. Each point represents the mean of three biological replicates, and the shaded areas indicate the standard deviation. Host range of TetV-2 The host range of T etV-2 was assessed across seven T etraselmis strains, including six T. striata isolates and the original T. striata BG strain from which T etV-2 was isolated (Fig. S9). Cultures were inoculated with purified T etV-2 and monitored after 13 d.p.i.. Cell concentrations and T etV-2 particle abundances were quantified by flow cytometry, while T sV DNA copy numbers were determined by qPCR. The tested T etraselmis strains displayed variable susceptibility to T etV-2, reflected in host mortality and viral productivity (Fig. 5). The highest mortality occurred in T. striata BG and RCC126, with cell densities reduced by 80–90% relative to uninfected controls at 13 d.p.i. In contrast, strains T etD, T etF, and LANL1001 displayed moderate preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint reductions in cell abundance (7–36%), while T. suecica exhibited enhanced growth in inoculated cultures (+39%). This latter result suggests that the material inoculated (including submicron cell debris not removed by the 0.45 μm prefiltration) may stimulate growth in this strain. Quantification of T sV-S2a revealed distinct strain-dependent patterns. T sV-S2a DNA was consistently detected across all T etraselmis strains, with copy numbers broadly correlating with T etV-2 production. Notably, T sV-S2a was detected in infected LANL1001 cultures despite the absence of T etV-2 replication, and low but reproducible T sV-S2a production was also observed in T. suecica, which did not support T etV-2 propagation. Control experiments excluded contamination or genomic integration as potential sources: (i) no amplification occurred when T. suecica genomic DNA was tested with T sV-S2a-specific primers, (ii) the T. suecica genome contained no sequences with significant similarity to primer targets, and (iii) no qPCR signal was detected in Conway medium inoculated with T etV-2 filtrates alone. These results suggest that T sV-S2a, present at initially undetectable levels in the T etV-2 inoculum, was able to replicate in the presence of T. suecica and T etV-2, despite the absence of productive T etV-2 infection. T sV-S3b was detected in most T etraselmis cultures, those of except T. suecica. In LANL1001, amplification was inconsistent: one replicate yielded no detectable signal, while the other two showed weak amplification close to the detection threshold, suggesting either stochastic presence or very low levels of the T sV-S3b target DNA. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint Figure 5:

1. Fistarol GO, Coutinho FH, Moreira APB, Venas T, Cánovas A, De Paula SEM, et al. Environmental and Sanitary Conditions of Guanabara Bay, Rio de Janeiro. Front Microbiol. 2015;6. doi:10.3389/fmicb.2015.01232 2. Jones JB, Rhodes LL. Suffocation of pilchards ( Sardinops sagax ) by a green microalgal bloom in Wellington Harbour, New Zealand. New Zealand Journal of Marine and Freshwater Research. 1994;28: 379–383. doi:10.1080/00288330.1994.9516627 Δ host abundance Δ virus abundance -50% 0% +50% -2e06 0 2e06 Ctl T. suecica LANL1001 TetF TetD RCC125 RCC126 T. striata BG Tsv-S2a TetV-2 (ml -1 ) 10 2 10 5 10 8 10 8 10 9 10 10 10 11 10 3 10 1 10 5 10 7 DNA copies.ml -1 Tsv-S3b Ctl T. suecica Inoculum LANL1001 TetF TetD RCC125 RCC126 T. striata BG A B preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint 3. Pizarro M. An Unusual and Massive Bloom of Tetraselmis Sp. in the Valparaiso Bay, Chile. OFOAJ. 2018;7. doi:10.19080/OFOAJ.2018.07.555717 4. Mata MT, Cameron H, Avalos V , Riquelme C. Identification and Characterization of a Novel Microalgal Strain from the Antofagasta Coast Tetraselmis marina AC16-MESO (Chlorophyta) for Biotechnological Applications. Plants. 2023;12: 3372. doi:10.3390/plants12193372 5. Hemaiswarya S, Raja R, Ravi Kumar R, Ganesan V , Anbazhagan C. Microalgae: a sustainable feed source for aquaculture. World J Microbiol Biotechnol. 2011;27: 1737–1746. doi:10.1007/s11274-010-0632-z 6. Fon-Sing S, Borowitzka MA. Isolation and screening of euryhaline Tetraselmis spp. suitable for large-scale outdoor culture in hypersaline media for biofuels. J Appl Phycol. 2016;28: 1– 14. doi:10.1007/s10811-015-0560-2 7. Steadman Tyler CR, Hovde BT, Daligault HE, Zhang XL, Kunde Y , Marrone BL, et al. High- Quality Draft Genome Sequence of the Green Alga Tetraselmis striata (Chlorophyta) Generated from PacBio Sequencing. Stajich JE, editor. Microbiol Resour Announc. 2019;8: e00780-19. doi:10.1128/MRA.00780-19 8. Keeling PJ, Burki F, Wilcox HM, Allam B, Allen EE, Amaral-Zettler LA, et al. The Marine Microbial Eukaryote Transcriptome Sequencing Project (MMETSP): illuminating the functional diversity of eukaryotic life in the oceans through transcriptome sequencing. PLoS Biol. 2014;12: e1001889. doi:10.1371/journal.pbio.1001889 9. Yao C, Ai J, Cao X, Xue S, Zhang W. Enhancing starch production of a marine green microalga Tetraselmis subcordiformis through nutrient limitation. Bioresource Technology. 2012;118: 438–444. doi:10.1016/j.biortech.2012.05.030 10. One Thousand Plant Transcriptomes Initiative. One thousand plant transcriptomes and the phylogenomics of green plants. Nature. 2019;574: 679–685. doi:10.1038/s41586-019-1693-2 11. Lim DKY , Schuhmann H, Thomas-Hall SR, Chan KCK, Wass TJ, Aguilera F, et al. RNA-Seq and metabolic flux analysis of Tetraselmis sp. M8 during nitrogen starvation reveals a two- stage lipid accumulation mechanism. Bioresource Technology. 2017;244: 1281–1293. doi:10.1016/j.biortech.2017.06.003 12. Lauritano C, De Luca D, Amoroso M, Benfatto S, Maestri S, Racioppi C, et al. New molecular insights on the response of the green alga Tetraselmis suecica to nitrogen starvation. Sci Rep. 2019;9: 3336. doi:10.1038/s41598-019-39860-5 13. Schvarcz CR, Steward GF. A giant virus infecting green algae encodes key fermentation genes. Virology. 2018;518: 423–433. doi:10.1016/j.virol.2018.03.010 14. Aylward FO, Abrahão JS, Brussaard CPD, Fischer MG, Moniruzzaman M, Ogata H, et al. Taxonomic update for giant viruses in the order Imitervirales (phylum Nucleocytoviricota). Arch Virol. 2023;168: 283. doi:10.1007/s00705-023-05906-3 15. Gajigan AP, Schvarcz CR, Conaco C, Edwards KF, Steward GF. Ultrastructural and transcriptional changes during a giant virus infection of a green alga. npj Viruses. 2025;3: 47. doi:10.1038/s44298-025-00128-7 16. Pagarete A, Grébert T, Stepanova O, Sandaa R-A, Bratbak G. Tsv-N1: A Novel DNA Algal Virus that Infects Tetraselmis striata. Viruses. 2015;7: 3937–3953. doi:10.3390/v7072806 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint 17. Roux S, Fischer MG, Hackl T, Katz LA, Schulz F, Yutin N. Updated Virophage Taxonomy and Distinction from Polinton-like Viruses. Biomolecules. 2023;13: 204. doi:10.3390/biom13020204 18. Kapitonov VV , Jurka J. Self-synthesizing DNA transposons in eukaryotes. Proc Natl Acad Sci USA. 2006;103: 4540–4545. doi:10.1073/pnas.0600833103 19. Pritham EJ, Putliwala T, Feschotte C. Mavericks, a novel class of giant transposable elements widespread in eukaryotes and related to DNA viruses. Gene. 2007;390: 3–17. doi:10.1016/j.gene.2006.08.008 20. Krupovic M, Bamford DH, Koonin EV . Conservation of major and minor jelly-roll capsid proteins in Polinton (Maverick) transposons suggests that they are bona fide viruses. Biol Direct. 2014;9: 6. doi:10.1186/1745-6150-9-6 21. Duponchel S, Fischer MG. Viva lavidaviruses! Five features of virophages that parasitize giant DNA viruses. Condit RC, editor. PLoS Pathog. 2019;15: e1007592. doi:10.1371/journal.ppat.1007592 22. Sizov DV , Polischuk VP. Cultivation, purification and crystallization of virus of green algae Tetraselmis viridis. Biopolym Cell. 2006;22: 243–245. doi:10.7124/bc.000737 23. Chase EE, Desnues C, Blanc G. Integrated viral elements suggest the dual lifestyle of Tetraselmis spp. Polinton-like viruses. Virus Evolution. 2022;8: veac068. doi:10.1093/ve/veac068 24. Roitman S, Rozenberg A, Lavy T, Brussaard CPD, Kleifeld O, Béjà O. Isolation and infection cycle of a polinton-like virus virophage in an abundant marine alga. Nat Microbiol. 2023;8: 332–346. doi:10.1038/s41564-022-01305-7 25. Santini S, Jeudy S, Bartoli J, Poirot O, Lescot M, Abergel C, et al. Genome of Phaeocystis globosa virus PgV-16T highlights the common ancestry of the largest known DNA viruses infecting eukaryotes. PNAS. 2013;110: 10800–10805. doi:10.1073/pnas.1303251110 26. Fischer MG, Hackl T. Host genome integration and giant virus-induced reactivation of the virophage mavirus. Nature. 2016;540: 288–291. doi:10.1038/nature20593 27. La Scola B, Desnues C, Pagnier I, Robert C, Barrassi L, Fournous G, et al. The virophage as a unique parasite of the giant mimivirus. Nature. 2008;455: 100–104. doi:10.1038/nature07218 28. Blanc G, Gallot-Lavallée L, Maumus F. Provirophages in the Bigelowiella genome bear testimony to past encounters with giant viruses. PNAS. 2015; 201506469. doi:10.1073/pnas.1506469112 29. Aylward FO, Abrahão JS, Brussaard CPD, Fischer MG, Moniruzzaman M, Ogata H, et al. Taxonomic update for giant viruses in the order Imitervirales (phylum Nucleocytoviricota). Arch Virol. 2023;168: 283. doi:10.1007/s00705-023-05906-3 30. Gilbert GE, Baleja JD. Membrane-binding peptide from the C2 domain of factor VIII forms an amphipathic structure as determined by NMR spectroscopy. Biochemistry. 1995;34: 3022– 3031. doi:10.1021/bi00009a033 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint 31. Krupovic M, Yutin N, Koonin EV . Fusion of a superfamily 1 helicase and an inactivated DNA polymerase is a signature of common evolutionary history of Polintons, polinton-like viruses, Tlr1 transposons and transpovirons. Virus Evol. 2016;2: vew019. doi:10.1093/ve/vew019 32. Redrejo-Rodríguez M, Ordóñez CD, Berjón-Otero M, Moreno-González J, Aparicio- Maldonado C, Forterre P, et al. Primer-Independent DNA Synthesis by a Family B DNA Polymerase from Self-Replicating Mobile Genetic Elements. Cell Reports. 2017;21: 1574– 1587. doi:10.1016/j.celrep.2017.10.039 33. Barth ZK, Hicklin I, Thézé J, Takatsuka J, Nakai M, Herniou EA, et al. Genomic analysis of hyperparasitic viruses associated with entomopoxviruses. Virus Evolution. 2024;10: veae051. doi:10.1093/ve/veae051 34. Burroughs AM, Zhang D, Schäffer DE, Iyer LM, Aravind L. Comparative genomic analyses reveal a vast, novel network of nucleotide-centric systems in biological conflicts, immunity and signaling. Nucleic Acids Res. 2015;43: 10633–10654. doi:10.1093/nar/gkv1267 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint Supplementary Tables T able S1 : Features of T etraselmis viruses Contig Length (bp) TIR length (bp) Genome conformation Number of protein genes Number of tRNA genes T etV-2 660 680 None Circular 640 4 T sv-S2a 19 063 1632 Linear 17 0 T sv-S2a.bis 19 170 1711 Linear 17 0 T sv-S2b 19 124 1724 Linear 18 0 T sv-S3b 18 038 583 Linear 17 0 T able S1 : Average nucleotide identity between T etraselmis viruses and their ETE guides Average Nucleotide Identity (ANI) C2208 (S2a) C1731 (S2b) C2184 (S3b) C0566 (S2a.bi s) Tsv-S2a Tsv- S2a.bis Tsv-S2b C1731 (S2b) 94,58 % C2184 (S3b) 65,31 % 62,03 % C0566 (S2a.bis) 98,53 % 96,05 % 65,81 % T sv-S2a 99,98 % 94,62 % 65,37 % 98,52 % T sv-S2a.bis 99,13 % 95,58 % 65,56 % 99,05 % 99,26 % T sv-S2b 94,78 % 99,81 % 62,06 % 96,01 % 94,79 % 95,72 % T sv-S3b 65,85 % 62,18 % 99,43 % 65,96 % 65,92 % 66,11 % 62,20 % preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint Supplementary Figures : Figure S1 | Endogenous TetV-like elements in Tetraselmis genomes. Left: Phylogenetic tree of viral DNA polymerases encoded by TetV-1, TetV-2, and TetV-like insertions in Tetraselmis genomes (data from Chase et al., 2018; DOI: 10.1093/ve/veac068). The TetV-1 DNA polymerase was used as a TBLASTN query to identify homologs in algal genomes. Protein sequences were aligned with MAFFT, and phylogenetic reconstruction was performed with IQ-TREE under default parameters. All branches received >80% ultrafast bootstrap support. Right: Genomic contigs encoding viral DNA polymerases. Two color-coded tracks indicate local GC content and the taxonomic affiliation of the best BLASTX protein match in GenBank NR. Colors: red, virus; green, green alga; blue, other eukaryotes; pink, prokaryotes. Viral insertion regions are underlined in red, with the average GC content of each insertion shown below. T. striata 2972 Pyramimonas orientalis virus Dishui Lake large algae virus 1 PMGC % T. striata 3242 PMGC % T. striata 784 PMGC % TetV-1 PMGC % T. chui C0378 Host average GC content: - T. chui : 53.0% - T. suecica : 53.0% - T. striata : 58.0% PMGC % T. suecica C5311 PMGC % T. suecica C6216 PMGC % TetV-2 PMGC % GC content scale 30% 70% 0 Kb 100 Kb 200 Kb 300 Kb 400 Kb 500 Kb 600 Kb 0.20 58.8% GC 63.6% GC 73.9% GC 71.0% GC 67.1% GC 66.0% GC 41.2% GC 57.8 % GC preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint Figure S2: phylogenetic trees of the transposases and transposase-associated proteins encoded by duplicated gene doublets. Phylogenetic reconstructions were generated with FastTree using default parameters. Branch support values were estimated with the SH-like local support method implemented in FastTree. Parasphingorhabdus sp. SCSIO 66989 WP 317080531.1 Sphingomonas pruni WP 157082394.1 Sphingomonas sp. PAMB00755 WP 270077527.1 Rubrivivax sp. MBC7939874.1 Alphaproteobacteria bacterium MBV8687861.1 Sphingomonas mali WP 157094128.1 Sphingomonas sp. 8AM WP 159760702.1 Sphingomonas hankookensis WP 272815506.1 Alphaproteobacteria bacterium HAK62192.1 Alphaproteobacteria bacterium MBV8535464.1 Sphingomonas deserti WP 199193114.1 Rhodospirillales bacterium MBK8210826.1 Hyphomonas sp. BRH c22 WP 299949975.1 Azospirillum brasilense WP 200529482.1 Chloroflexota bacterium MBC8496267.1 bacterium MCY4435134.1 Halomonas xianhensis WP 092842674.1 Ktedonobacter racemifer WP 007918235.1 Rickettsia endosymbiont of Culicoides newsteadi WP 239832580.1 Leptolyngbya sp. PCC 6406 WP 027268853.1 filamentous cyanobacterium CCT1 PSN07920.1 Petrachloros mirabilis WP 161825414.1 Geminocystis herdmanii WP 144051420.1 Acaryochloris thomasi WP 110984260.1 Candolleomyces aberdarensis RXW12687.1 Dictyocoela muelleri KAG0438599.1 Hydra vulgaris XP 047144825.1 Chitinophagia bacterium NBO50101.1 Flavobacteriaceae bacterium NBW28744.1 Spirochaetia bacterium NBK24064.1 Alphaproteobacteria bacterium NDB85771.1 Tetraselmis virus 1 YP_010783652.1 (TetV_624) TetV2 00616 TetV2 00269 TetV2 00420 TetV2 00601 TetV2 00636 TetV2 00126 TetV2 00202 94 98 83 0 100 89 96 63 57 99 98 92 85 95 98 99 63 92 93 100 18 94 99 71 91 100 63 38 99 81 84 96 1.00 uncultured virus CAH6420238.1 Bacteroidota bacterium NCG05012.1 Gaertneriomyces sp. JEL0708 KAJ3176077.1 Coccomyxa sp. Obi BDA45239.1 Klebsormidium nitens GAQ89740.1 Lipomyces tetrasporus XP 056042647.1 Mucor circinelloides 1006PhL EPB90321.1 Rhizopus microsporus KAG1180579.1 Rhizopus microsporus CEG63517.1 Rhizopus microsporus ORE12514.1 Thamnidium elegans KAI8050113.1 Cokeromyces recurvatus XP 051384016.1 Shrimp white spot syndrome virus NP 478008.1 Dishui Lake large algae virus 1 QIG60107.1 Tetraselmis virus 1 YP_010783651.1 (TetV_623) Chlamydomonas sp. UWO 241 KAG1669784.1 T etradesmus obliquus WIA10818.1 Haematococcus lacustris KAJ9510933.1 Haematococcus lacustris KAJ9528213.1 Haematococcus lacustris KAJ9530676.1 Haematococcus lacustris KAJ9516151.1 Haematococcus lacustris KAJ9507663.1 Haematococcus lacustris KAJ9508855.1 Haematococcus lacustris KAJ9517900.1 Haematococcus lacustris KAJ9518757.1 Haematococcus lacustris KAJ9509524.1 Haematococcus lacustris KAJ9526513.1 Haematococcus lacustris KAJ9528516.1 TetV2 00637 TetV2 00617 TetV2 00600 TetV2 00419 TetV2 00268 TetV2 00203 TetV2 00125 TetV2 00239 0 97 87 100 86 100 94 33 99 99 17 79 23 16 70 92 99 100 97 96 85 58 86 98 100 93 80 26 25 80 0.50 Transposase Transposase- associated protein Figure SX: phylogenetic trees of the transposases and transposase-associated proteins encoded by the duplicated gene doublets. Phylogenetic reconstructions were done using FastTree and default parameters. Branch supports were calculated using the SH-like local supports methods as implemented in FastTree. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint Figure S3 : Regions of the LANL1001 genome covered by Illumina short reads. Each panel shows position-specific read coverage (orange, y-axis). Below the coverage plots are the corresponding regions assembled from short reads (thick blue lines) and long reads (thick green lines), together with annotations of endogenous TetV-like elements (ETEs), including genes (colored arrows) and terminal inverted repeats (TIRs, thick red lines). Gene color codes are indicated in the figure. For each region, the region identifier, contig GenBank accession number, and genomic coordinates within the contig are provided. F i g S X : R e g i o n s o f t h e L A N L 1 0 0 1 g e n o m e c o v e r e d b y I l l u m i n a s h o r t r e a d s . E a c h r e g i o n i s r e p r e s e n t e d b y a g raph ( o r a n g e ) s h o w i n g t h e p o s i t i o n s p e c i fi c r e a d c o v e r a g e ( y - a x i s ) . B e l o w t h e g r a p h s a r e s h o w n r e g i o n s c o v e r e d b y contigs o f t h e i n i t i a l s h o r t - r e a d a s s e m b l y ( t h i c k b l u e h o r i z o n t a l l i n e s ) a n d l o n g - r e a d a s s e m b l y ( t h i c k g reen h o r i z o n t a l l i n e s ) a nd a n n o t a t i o n o f t h e E T E s , i n c l u d i n g g e n e s ( c o l o r e d a r r o w s ) a n d T I R s ( t h i c k r e d l i n e s ) . C olor c o d e f o r g e n e s i s s hown. Region id, contig Genbank accession number and positions in the contig are given beside each graph. ATPase DNA_pol_B Helicase-Primase-D5 mCP MCP ProtKinase Tyrosine recombinase TVSG_00023 TVSG_00024 C3212 C0552 C0694 C0826 C0942 C1307 C3296 C0674 C1258 C0668 10 100 1000 10000 10 100 1000 10000 10 100 1000 10000 10 100 1000 10000 10 100 1000 10000 10 100 1000 10000 10 100 1000 10000 10 100 1000 10000 10 100 1000 10000 10 100 1000 10000 C0566 10 100 1000 10000 C0660 10 100 1000 10000 C1731 10 100 1000 10000 C2184 10 100 1000 10000 C2208 10 100 1000 10000 VCJN01002215 (8093-27186) VCJN01001737 (40388-64980) VCJN01000569 (3066-20976) VCJN01002191 (54413-72662) VCJN01000555 (111979-128973) VCJN01000664 (44505-61458) VCJN01001313 (56068-78553) VCJN01003222 (611867-632801) VCJN01003307 (4-19958) VCJN01000678 (2096-19834) VCJN01000698 (21909-34058) VCJN01000672 (11613-22054) VCJN01000830 (2114-12546) VCJN01000946 (46240-54446) VCJN01001263 (108836-113748) Tsv-S2a Tsv-S2a bis Tsv-S2b Tsv-S3b Tsv-S3a Sequencing depth Tandem duplication preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint Figure S4: Minimum evolution phylogenetic trees of RBCL (A) and 18S rDNA (B) sequences generated using MEGA X. Figure S5 : Sequence alignments at insertion sites between Tsv genomes and ETE guides. VCJN01001471 T. striata LANL10001U05039.1 T etraselmis convolutaeJQ315728.1 T etraselmis sp. KMMCC40JQ315807.1 T etraselmis sp. KMMCC332U41900.1 T etraselmis sp. RG-07JQ315813.1 T etraselmis striata MMCC1157JQ315812.1 T etraselmis sp. KMMCC1609X70802.1 T. striataJN904000.1 Tetraselmis striata SAG41.85OQ220339.1 T etraselmis striata BEA0097BKJ756817.1 T etraselmis sp. CCAP66/15MG022700.1 T etraselmis sp. CCAP66/52T. striata BGOQ220340.1 T etraselmis striata BEA0098FJ559402.1 T etraselmis striata KMMCC P-37OQ687072.1 T etraselmis sp. BEA0651BOQ687071.1 T etraselmis striata BEA0650BOQ687074.1 T etraselmis striata BEA0653BOQ687073.1 T etraselmis striata BEA0652BGQ917218.1 T etraselmis striata strain KMMCC P-56JQ315733.1 T etraselmis sp. KMMCC 258JQ315731.1 T etraselmis sp. KMMCC 256JQ315730.1 T etraselmis sp. KMMCC 231GQ917216.1 T etraselmis sp. KMMCC P-60GQ917214.1 T etraselmis sp. KMMCC P-55JQ315732.1 T etraselmis sp. KMMCC 257JQ315734.1 T etraselmis sp.GQ917219.1 T etraselmis sp. KMMCC P-63GQ917217.1 T etraselmis sp. KMMCC P-62JQ315729.1 T etraselmis sp. KMMCC 171FJ559390.1 T etraselmis sp. KMMCC P-21FJ559385.1 T etraselmis sp. KMMCC P-13FJ559378.1 T etraselmis striata KMMCC P-5MG022703.1 T etraselmis sp. CCAP 66/73FR744745.1 T etraselmis sp. M2GQ917215.1 T etraselmis sp. KMMCC P-59FJ559400.1 T etraselmis sp. KMMCC P-42FJ559393.1 T etraselmis sp. KMMCC P-17FJ559397.1 T etraselmis sp. KMMCC P-33FJ559407.1 T etraselmis sp. KMMCC P-52FJ559388.1 T etraselmis sp. KMMCC P-24FJ559403.1 T etraselmis striata KMMCC P-43KT860869.1 T etraselmis striata RCC131KX109779.1 T etraselmis striata WT3OQ687067.1 T etraselmis sp. BEA0077BOQ687066.1 T etraselmis sp. BEA0076BFR744761.1 T etraselmis sp. M18KT860858.1 T etraselmis sp. RCC120FJ559405.1 T etraselmis sp. KMMCC P-48FR744744.1 T etraselmis sp. M1KT860867.1 T etraselmis 'chuii' RCC129KT860864.1 T etraselmis sp. RCC126FR744759.1 T etraselmis sp. M16GQ917221.1 T etraselmis sp. KMMCC P-57FJ559394.1 T etraselmis sp. KMMCC P-31MG022691.1 T etraselmis inconspicua CCAP 66/19BKJ756818.1 T etraselmis inconspicua CCAP 66/19CMG022697.1 T etraselmis sp. CCAP 66/29MG022696.1 T etraselmis sp. CCAP 66/26MG022695.1 T etraselmis sp. CCAP 66/25JQ315802.1 T etraselmis sp. KMMCC 106MG022699.1 T etraselmis sp. CCAP 66/50KU561160.1 T. subcordiformis Xmm14S3KU561142.1 T. subcordiformis Xmm14S2KU561107.1 T. subcordiformisFJ559384.1 T etraselmis 'carteriiformis' KMMCC P-12FJ559380.1 T etraselmis subcordiformis KMMCC P-6KT860862.1 T etraselmis sp. RCC124MT489380.1 T etraselmis sp. SMS19MN721295.1 T etraselmis tetrathele CCAP 66/41MG022702.1 T etraselmis sp. CCAP 66/72MG022701.1 T etraselmis sp. CCAP 66/64JQ315803.1 T etraselmis sp. KMMCC 139JQ315801.1 T etraselmis sp. KMMCC 84DQ207405.1 T etraselmis chuiiJN903999.1 T etraselmis chuii SAG 1.96EF473736.1 T etraselmis sp. T sbreJQ315810.1 T etraselmis sp. KMMCC 1156MT489359.1 T etraselmis sp. SMS19MH166729.1 T etraselmis sp. UPMC-A0074FJ559379.1 T etraselmis sp. KMMCC P-8MG022693.1 T etraselmis chuii CCAP 66/21BFJ517749.1 T etraselmis tetrathele KMMCC P-2FJ517748.1 T etraselmis subcordiformis KMMCC P-1FR744760.1 T etraselmis sp.MN720749.1 T etraselmis tetrathele CCAP 66/1AFR744762.1 T etraselmis sp. T1FR744746.1 T etraselmis sp. M3FR744749.1 T etraselmis sp. M6AY954898.1 T etraselmis sp. TEQL01FJ559376.1 T etraselmis sp. KMMCC P-3FJ559406.1 T etraselmis sp. KMMCC P-49MW195044.1 T etraselmis chuii TCBG-2MW195043.1 T etraselmis chuii TCBG-1FJ559381.1 T etraselmis suecica KMMCC P-9KT860859.1 T etraselmis sp. RCC121HF931098.1 T etraselmis chuiiKJ756816.1 T etraselmis gracilis CCAP 66/13AY954899.1 T etraselmis sp. NT18 MZ435989.1 T. jejuensis YO35MZ435987.1 T. jejuensis YO31MZ435988.1 T. jejuensis YO32 0.0020 T. striata T. striata KU167097 T etraselmis sp. CCMP 881 EU555175 T etraselmissp. CCAP66/4. MZ936271 T etraselmis sp. SC-5 T. striata BG VCJN01001457 T. striata LANL1001 AY954897 T etraselmis sp. NT18 NC 067041 T etraselmis suecica DQ173247 T etraselmis suecica CCMP 904 NC 029807 Scherffelia dubia NC 065695 T etraselmis marina OQ942923 Picochlorum sp. BH-2019 KX232643 T etrabaena socialis NC 029673 Chlorotetraedron incus NC 029671 Mychonastes homosphaera NC 029040 Ulva fasciata NC 042255 Ulva lactuca AP018696 Ulva ohnoi MZ561583 Ulva laetevirens NC 077591 Ulva meridionalis NC 067776 Ulva tepida MN853879 Ulva prolifera 0.01 A B RBCL 18S Fig SX: Minimum evolution phylogenetic trees of RBCL (A) and 18S rDNA (B) se q u e n c e s preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint Figure S6 : Alignment of assembled Tetraselmis virus (Tsv) genomes against endogenous Tsv element (ETE) reference regions used for Illumina short-read recruitment. Pairwise alignments were generated using Dotter from the SeqTools package (Sonnhammer & Durbin, 1995; Gene 167: GC1–10; doi:10.1016/0378-1119(95)00714-8). T. striata LANL1001 C2208 (bp) T. striata LANL1001 C0566 (bp) Tsv-S2a Tsv-S2a.bis Tsv-S2b Tsv-S3b 2K 4K 6K 8K 10K 12K 14K 16K 18K T. striata LANL1001 C1731 (bp) T. striata LANL1001 C2184 (bp) 2K 4K 6K 8K 10K 12K 14K 16K 18K 20K 22K 24K 2K 4K 6K 8K 10K 12K 14K 16K 18K 2K 4K 6K 8K 10K 12K 14K 16K Fig. S5. Alignment of assembled T sv genomes against endogenous T sv element (ETE) regions used as references for Illumina short-read recruitment. Alignments were generated with the program Dotter from the SeqT ools package (DOI: 10.1016/0378-1119(95)00714-8). preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint Figure S7. Transmission electron microscopy of Tetraselmis viruses during infection of Tetraselmis striata (TS). (A1–C2) Ultra-thin sections of infected cells at 24 hours post-infection (hpi). Panels A1, B1, and C1 show overviews of cellular organization; panels A2, B2, and C2 display higher magnifications of preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint the boxed regions. In A2 and B2, high-magnification views reveal assembly of giant virus particles within putative viral factories (VF), with membrane fragments (arrowheads), as well as empty and partially filled particles. In C2, smaller Tetraselmis virus-like particles (Tsv) are concentrated (dashed line) at the cell periphery. Panels A2a and A2b show negatively stained mature giant virus particles at the same scale as A2. (D1–F2) Ultra-thin sections of infected TS cells at 48 hpi. D1, E1, and F1 show cellular overviews, while D2, E2, and F2 depict corresponding higher-magnification views. D2 and E2 reveal mature and immature empty giant virus particles, whereas F2 shows Tsv particles concentrated in a central region (dashed line). (G1–I2) Ultra-thin sections of infected TS cells at 72 hpi. G1, H1, and I1 show overviews of cellular ultrastructure, with G2, H2, and I2 providing higher-magnification views of the boxed areas. Most giant virus particles appear empty (G2). Peripheral and central regions rich in Tsv particles (dashed lines in H2 and I2) are evident, with Tsv assembling into pseudo-crystalline arrays. Abbreviations: NC, nucleus; NCL, nucleolus; IB, inclusion body; CHL, chloroplast; PYR, pyrenoid. Figure S8. Quantification of TsV DNA copy numbers at 0 and 13 days post infection (dpi). TsV abundance was determined by quantitative PCR (qPCR) using strain-specific primer sets. The inoculum consisted of a TsV mixture dominated by strain TsV-S2b. Bars represent the mean values of three biological replicates, and error bars indicate the standard deviation. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint Figure S9. 18S rRNA phylogenetic relationships among Tetraselmis strains used in the host range assay. Strains tested in this study are shown in bold. TetV-2 replication levels are represented by stars, where an increasing number of stars denotes higher viral replication. T etraselmis sp. permissive to T etV-1|MH055449.1 X68484|Scherffelia dubia T etraselmis sp. permissive to T etV-1|MH055453.1 T etraselmis sp. permissive to T etV-1|MH055456.1 T etraselmis sp. permissive to T etV-1|MH055454.1 JN376804|T etraselmis astigmatica AJ431370|T etraselmis kochiensis KT860913|T etraselmis convolutae HE610130|T etraselmis cordiformis HQ651184|T etraselmis indica T etraselmis sp. permissive to T etV-1|MH055452.1 T etraselmis sp. permissive to T etV-1|MH055445.1 T etraselmis sp. permissive to T etV-1|MH055448.1 T etraselmis sp. permissive to T etV-1|MH055444.1 KY054995|T etraselmis marina Prasinocladus sp. malaysianus KT860871|T etraselmis rubens MZ435987|T etraselmis jejuensis JN903999|T etraselmis chuii T. suecica CCMP904 JF489950|T etraselmis suecica FJ517748|T etraselmis subcordiformis FJ559384|T etraselmis carteriiformis T.striata LANL1001 TetF T. striata BG TetD RCC125.KT860863.1 RCC126.KT860864.1 X70802|T etraselmis striata 0 92 97 6 96 99 99 98 75 99 17 73 99 74 98 75 97 94 99 85 89 97 94 48 94 41 33 0.02 Strains with detected replication of TetV-2 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint Supplementary Data : Sequences of Tetraselmis striata strain GB gene markers : >RBCL-T.striata.sp.BG GCGTACCCTTTAGATTYATTTGAAGAAGGTTCAGTTACAAACTTATTTACTTCTATTGTAG KTAACGTTTTTGGTTTCAAAGCTTTACGTGCTCTTCGTTTAGAAGATTTACGTATTCCAGT TGCTTACTGTAAAACTTTTACAGGTGCTCCGCACGGTATTCAAGTTGAACGTGATAAATT AAACAAATATGGACGTGGTTTATTAGGTTGTACTATTAAACCAAAATTAGGTCTTTCTGCT AAAAACTACGGTCGTGCTTGTTACGAATGTTTACGTGGTGGTTTAGATTTTACAAAAGAT GATGAAAACGTAAACTCACAAGCATTTATGCGTTGGAGAGATCGTTTCCTATTCGTATCA GAAGCTATCTACAAATCACAAGCAGAAACTGGTGAAATTAAAGGTCACTACTTAAACGT AACAGCAGGTACTTGTGAAGAGATGATGAAGCGTGCTGAATGTGCCGCAGGTTTCGGTG TACCAATTGTTATGCACGATTACTTAACAGGTGGTTTTACAGCTAACACTTCATTAGCTAT TTACTGTCGTGATAATGGTTTACTATTACACATTCACCGTGCTATGCACGCAGTTATTGAC CGTCAACGTAATCACGGAATTCACTTCCGTGTTTTAGCAAAAGCTTTACGTATGTCTGGT GGGGATCACCTTCACTCAGGTACTGTTGTAGGTAAATTAGAAGGTGAACGTGAAGTTAC TTTAGGTTTCGTAGATTTAATGCGTGATGCTTACGTAGAAAAAGATCGTTCTCGTGGT >18S-T.striata.sp.BG aatcatgataacttcacgaatcgcatggcctccgcgccggcgatgtttcattcaaatttctgccctatcaatttgcgatggtaggatagaggcctac catggtggtaacgggtgacggaGaattagggttcgattccggagagggagcctgagaaacggctaccacatccaaggaaggcagcaggcg cgcaaattacccaatcctgacacagggaggtagtgacaataaataacaataccgggcttttcaagtctggtaattggaatgagtacaatctaaaca accttaacgaggatccattggagggcaagtctggtgccagcagccgcggtaattccagctccaatagcgtatatttaagttgctgcagttaaaaa gctcgtagttggatttcggatgggatttgccggtccgccgtttcggtgtgcactggccagtcccatcttgttgtcggggactagctcctgggcttca ctgtccgggactaggagctgacgaggttactttgagtaaattagagtgttcaaagcaagcctacgctctgaatacattagcatggaataacatgat aggactctggcttatcttgttggtctgtgagaccagagtaatgattaagagggacagtcgggggcattcgtatttcattgtcagaggtgaaattcttg gatttatgaaagacgaacttctgcgaaagcatttgtcaaggatgttttcattaatcaagaacgaaagttgggggctcgaagacgattagataccgtc ctagtctcaaccataaacgatgccgactagggattggcagacgtttttttgatgactctgccagcaccttatgagaaatcaaagtttttgggttccgg ggggagtatggtcgcaaggctgaaacttaaaggaattgacggaagggcaccaccaggcgtggagcctgcggcttaatttgactcaacacggg aaaacttaccaggtccagacatagtgaggattgacagattgagagctctttcttgattctatgggtggtggtgcatggccgttcttagttggtgggtt gccttgtcaggttgattccggtaacgaacgagacctcagcctgctaaatagttactcctactttggtaggaggtgaacttcttagagggactattgg cgtttagccaatggaagtgtgaggcaataacaggtctgtgatgcccttagatgttctgggccgcacgcgcgctacactgatgcattcaacgagcc tagccttgaccgagaggtccgggtaatctttgaaactgcatcgtgatggggctagattattgcaattattaatcttcaacgaggaatgcctagtaag cgtgattcatcagatcgcgttgattacgtccctgccctttgtacacaccgcccgtcgctcctaccgattgaatgtgttggtgaggagttcggattggc agtttgtggtggttcgccactgcttacagctgagaagttctccaaaccgccccatttagaggaaggagaagtcgtaacaaggtttccgtaggtgaa cctgc preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted October 13, 2025. ; https://doi.org/10.1101/2025.10.09.676808doi: bioRxiv preprint