Endogenous DNA viruses take center stage in eukaryotic genome evolution

Endogenous viral elements (EVEs) are common components of eukaryotic genomes that play key roles in transcriptional regulation, antiviral defense, and genetic disorders (1, 2). It has long been recognized that retroviruses often comprise a large bulk of these EVEs, but discoveries over the last few decades have highlighted that nonretroviral RNA viruses and single-stranded DNA viruses are also prevalent components of eukaryotic genomes (3). Recent studies have begun to show that many EVEs also derive from the complex genomes of large double-stranded DNA viruses (4–6), demonstrating that all major viral lineages can directly shape eukaryotic genomes in this way. In this issue, Bellas et al. elegantly demonstrate that polinton-like viruses (PLVs) are ubiquitous EVEs in the genomes of unicellular eukaryotes (protists), further highlighting the extensive role of dsDNA EVEs in eukaryotic genome evolution (7).

Bellas et al. conducted this analysis by developing a bioinformatic pipeline specifically for the detection of PLVs, a group of viruses related to Maverick–Polinton elements. Although Maverick–Polintons were once thought to be transposons, recent detection of major capsid proteins encoded by these elements suggests that they are themselves bona fide endogenous viruses. The evolutionary distance between different groups of PLVs is considerable, and Bellas et al. astutely note that previous studies have likely overlooked their presence in diverse eukaryotic genomes due to their high divergence from characterized references. Another key element to the success of this approach is the recent use of long-read sequencing in many protist genome projects. Indeed, the authors found that some protist genomes encode an enormous diversity of PLVs, including many nearly-identical copies. Traditional short-read sequencing approaches are unable to resolve these highly repetitive elements, often leading to their absence in many draft genome assemblies.

Although Bellas et al. focus their analysis on protists, the significance of their findings encompasses a broader range of hosts due to the presence of many PLVs in diverse animal genomes.

The presence of a vast diversity of PLVs in many protist genomes raises further questions regarding their mode of proliferation. Are all PLVs strictly viruses that integrate as part of their infectious cycle? Or can some proliferate in a transposon-like manner within a genome? The broad range of distinct PLVs in individual protist genomes suggests that much of this diversity can be attributed to viral endogenization, but it is still likely that some PLVs span a continuum of propagation modalities (Fig. 1), similar to the “dual life cycle” seen in many retroviruses (8, 9). Some transposons and other selfish genetic elements may even have evolved from PLVs through loss of the morphogenetic module, indicating that PLVs are a kind of wellspring of eukaryotic transposable elements (8). This appears to be an evolutionary trajectory taken by other DNA viruses as well; the remarkable Teratorn transposons have been proposed to form through fusion of an alloherpesvirus and piggyBac transposons, leading to “giant transposons” that are prevalent in many fish species (10).

Fig. 1.

A schematic of known life strategies of PLVs. Three distinct aspects are shown. 1. Virophages coinfect hosts along with giant viruses and replicate within giant virus factories. 2. PLVs sometime can independently infect a host and replicate within host nucleus. 3. In both life strategies, DNA of these viruses can integrate within host genome and possibly propagate in the case of some hosts through duplication. PLV - polinton-like virus, VF - virus factory, NUC - nucleus, MT - mitochondria.

Regardless of their mode of transmission, what is the impact of these PLVs on their eukaryotic hosts? There are several clues from recent experimental work. A related group of dsDNA viruses—virophages—can integrate into host genomes and act as a kind of inducible antiviral defense against coinfecting giant viruses. A hallmark study of the heterotrophic flagellate Cafeteria roenbergensis demonstrated that an endogenous virophage reactivates upon infection by a giant virus, parasitizing the virus factories of the latter and suppressing giant virus replication in future rounds of infection (11). Subsequent long-read sequencing revealed that dozens of virophage-like elements are distributed unevenly across the genomes of C. roenbergensis field isolates (12), suggesting that their presence may be driven by a dynamic tripartite coevolution between host, giant virus, and virophage. Intriguingly, a similar phenomenon was recently reported for a PLV associated with the abundant marine haptophyte Phaeocystis globosa (13), demonstrating that these multipartite viral interactions are widespread in the environment and explain the distribution of many PLVs in protist genomes.

It may be tempting to believe that most endogenous PLVs are hyperparasites of giant viruses, but many of these EVEs are likely the result of autonomous viral propagation. Many viral diversity surveys have shown that free PLVs are widespread in various environments (14, 15), and at least one PLV that infects the green alga Tetraselmis striata (TsV1) appears capable of independently infecting and lysing its host (16). Endogenous viruses closely related to TsV1 have been identified in Tetraselmis genomes (17), suggesting that PLVs also exist as autonomous viruses independent of giant virus infection. Indeed, latency is often a successful strategy that is commonly employed by dsDNA viruses of bacteria, archaea, and animals, and there is little reason to doubt that some PLVs can dynamically move between lysis/latency in a similar manner to other viruses. It is plausible that many PLVs use latency as a strategy to persist while host populations are low or growth conditions are not suitable for viral replication.

There is no a priori reason to believe that hyperparasitism of giant viruses and autonomous viral proliferation are always mutually exclusive, however. It is plausible that some PLVs or virophages may utilize both strategies depending on prevailing environmental conditions and giant virus abundances. Just as many symbionts can exploit their hosts in some situations, some antiviral PLVs may sometimes proliferate independently or evolve into autonomous viruses. We are only recently beginning to grasp the extent of virus–virus interactions in the biosphere, and it seems likely that the tripartite host–virophage–giant virus dynamic is part of a broader evolutionary milieu that includes a wide range of complex viral infection strategies. Recent studies have suggested that coinfections are common in the environment (18), underscoring the need to advance our understanding of how this influences cellular outcomes and viral evolution.

Although Bellas et al. focus their analysis on protists, the significance of their findings encompasses a broader range of hosts due to the presence of many PLVs in diverse animal genomes. Indeed, a recent study examining Maverick–Polinton groups in vertebrate genomes found that they are widespread, often encode morphogenesis modules consistent with their existence as EVEs, and in some cases are undergoing purifying selection indicative of a possible beneficial role to the host (19). Another study reported the widespread presence of another PLV group (adintoviruses, now classified within the virus class Polintoviricetes) in the genomes of many invertebrates (20). Clearly, regardless of their precise role, PLVs are distributed in a wide range of genomes across the eukaryotic tree of life.

All of this serves to highlight the emerging richness of endogenous DNA viruses in eukaryotic genomes. Other studies have found many large eukaryotic EVEs derived from giant viruses (phylum Nucleocytoviricota) (4, 6), a group known for its complex genomes that often encode numerous metabolic enzymes, complex DNA repair mechanisms, and even key proteins involved in cytoskeletal structure and DNA packaging (21–23). Analysis of gene transfer between viruses and eukaryotes has shown that many eukaryotic lineages have acquired a diverse range of genes from DNA viruses throughout their evolutionary history, underscoring the important long-term impacts that viral endogenization can have on cellular evolution (24). These studies put endogenous DNA viruses at center stage for future work on eukaryotic genomics and evolution.