Roles of viruses in the environment


Viruses are important microbial predators that influence global biogeochemical cycles and drive microbial evolution, although their impact is often under appreciated. Viruses reproduce after attaching and transferring their genetic material into a host cell. The host's cellular machinery is then redirected to the making of more viruses and results in the death of the host cell in the vast majority of cases. Viruses have developed intriguing mechanisms to utilize host proteins for their own defence and for shifting metabolism from host to virus, a topic that is reviewed for bacterial viruses in this special issue of Environmental Microbiology (Roucourt and Lavigne, 2009).

Globally, there are an estimated 1e31 virus-like particles. Currently, it is thought that most of the viruses are phages that infect bacteria, but archaeal and eukaryotic viruses are certainly important components of most ecosystems. Since the average half-life of free viruses in most ecosystems is ∼48 h, an estimated 1e27 viruses are produced every minute. This means that roughly 1e25 microbes, or about 100 million metric tons, die every 60 s due to viruses. Viral predation, in combination with protist grazing, is able to maintain microbial numbers at values less then the carrying capacity of the system and as such plays an important role in controlling microbial communities. In this special issue Sandaa and colleagues (2009) report on the simultaneous control of microbial diversity by both viral predation (top-down control) and substrate availability (bottom-up control).

Much of our knowledge about the roles of viruses in natural environments comes from studies of marine microbial communities. In the world's oceans, about half of the organic matter produced by photosynthesis supports the production of new heterotrophic microbes (both Bacteria and Archaea). Viruses and protists then kill roughly equal proportions of these (Fuhrman and Noble, 1995). The lysed cells become dissolved organic matter which can be used by other heterotrophic bacteria. This means that viral-mediated mortality increases net respiration, the release of CO2 and nutrient recycling in the world's oceans. Viruses and their microbial prey are also extremely diverse, abundant and active in marine sediments (Danovaro et al., 2002; Breitbart et al., 2004a). Moreover, viruses affect primary productivity by killing diatoms, dinoflagellates and cyanobacteria, as well as by releasing nutrients (Suttle et al., 1990). Thus, viruses can account for a very significant part of the ocean's carbon cycling.

Viruses seem to be ubiquitous and have been reported from any environment where life is present, from freshwaters to the sands of the Sahara desert. However, very little is known of the ecological roles of viruses in most ecosystems. In soils, potentially the biggest biosphere on the planet, most viral studies have concentrated on estimating their abundance and taxonomy. Viruses are associated with the rhizosphere of plants (Ackermann, 1997) and are also common in some of the harshest environments in the planet, ranging from hot springs (Rice et al., 2001; Rachel et al., 2002; Breitbart et al., 2004b; Redder et al., 2009) to hypersaline waters (Nuttall and Dyall-Smith, 1993). In many of these extreme ecosystems, viruses are the only known microbial predators.

Metagenomic analyses of natural communities and man-created niches have shown that viruses are extremely diverse and novel. For example, 1 kg of marine sediment may contain over a million different viral types and 200 l of seawater may contain about 5000 viral types (Breitbart et al., 2002; 2004a). And there may be at least 1000 different viruses living in the human gut (Breitbart et al., 2003). The vast majority (> 70%) of genetic material carried by these viruses is completely uncharacterized and natural viral communities probably represent the largest unexplored area of genetic information space left on the planet. Metagenomics analyses of blood have also shown that healthy humans carry a number of novel, unknown viruses (Breitbart and Rohwer, 2005), including phage to known human pathogens (Gaidelytëet al., 2007). The impact of these viruses on human health is currently unknown, but we expect an explosion of research in this area over the next couple of years.

In addition to their influences on biogeochemical cycles, viruses drive microbial evolution by natural selection for microbes resistant to infection and via lateral gene transfer. Many viruses are strain-specific predators. Therefore as a particular microbial strain becomes dominate in a system, its viral predators will expand exponentially and kill it off. This will leave a niche for another microbial strain to grow into, which will be subsequently killed off by another viral type. This means that the dominant microbial species within a system will be constantly turned over. This ‘kill-the-winner’ hypothesis may explain much of the observed microbial diversity and changes in community structure (Thingstad, 2000).

Viruses are also important exchangers of genetic information between hosts, because they inject their genomes into the host cells. For example, most of the completely sequenced microbial genomes contain proviral sequences. Proviruses are viruses that have integrated their genomes into the host's genome and are replicated with the host. Most proviruses can become active at a later date and subsequently end up killing their host. Many proviruses also express genes, which can dramatically alter the phenotype of the host cell. The majority of environmental strains of Vibrio cholerae, for example, are not human pathogens until they are infected with a provirus carrying the cholera toxin. Acquisition and loss of proviruses is one of the most common mechanisms of lateral gene transfer.

Viruses also move ecologically important genes from host to host. For example, viruses that infect the marine cyanobacteria Prochlorococcus and Synechoccus often carry a gene (psbA) that encodes a protein central to photosynthesis (Mann et al., 2003; Lindell et al., 2004). These viruses express this and other photosynthesis genes during infection and are thought to use photosynthetic proteins to keep that host cell alive and produce energy during the infection cycle (Lindell et al., 2005; Clokie et al., 2006). However, in some cases, when the virus moves this central gene from one microbial strain to another, a recombination event can incorporate parts of the viral gene into the host (Zeidner et al., 2005; Sullivan et al., 2006). This event may simultaneously inactivate the viral genome, enabling the host to survive infection, and change the genotype of the host. Another intriguing example is of the horizontal transfer of the ceramide-producing sphingolipid biosynthesis genes from coccolithophores to their viruses (Wilson et al., 2005; Monier et al., 2009). These genes, which may be involved in triggering programmed cell death, are expressed during infection (Allen et al., 2006) as well as during a coccolithophore bloom in a natural mesocosm microbial community (Pagarete et al., 2009). These examples, where viruses have maintained host genes and actively express them during infection, indicate that viruses are not merely vectors for horizontal gene transfer but that gene transfer impacts the evolution of both microbial hosts and viruses.

Viruses not only move genes from one organism to another, they are also able to move genetic material between ecosystems. Some viral sequences have been found to be ubiquitously spread through the biosphere (Breitbart et al., 2004c; Short and Suttle, 2005). There is also evidence that viruses from one environment can successfully infect and replicate on microbes from unrelated environments (Wilhelm and Matteson, 2008). These results provide support that viruses can move throughout the world and move genes between ecosystems. Similarly a recent study of RNA viruses in human stool samples showed that plant viruses are efficiently passing though the human gut and are disseminated with the seeds (Zhang et al., 2006). In this way, viruses can use animals to move from place to place.

Our knowledge of environmental viruses has increased greatly over the last decade, yet we still have much to learn about even the best-studied environments as the reports in this special issue show. New viruses are still being discovered and their genomes are continuing to reveal a multitude of novel genes and new potential functions (Redder et al., 2009; Sullivan et al., 2009). Enormous viral diversity is being uncovered in every new habitat investigated (Rosario et al., 2009) and we are just beginning to understand the temporal and spatial variability of the diversity of natural viral populations (Chen et al., 2009; Short and Short, 2009). Seasonal and diel changes in viral production are also reported (Winget and Wommack, 2009) and virus-mediated bacterial mortality is shown to significantly impact carbon cycling in the Southern Ocean (Evans et al., 2009).

The extent of microbial mortality is dependent on their susceptibility to viral infection, yet many microbes are resistant to infection. Varying degrees of resistance and susceptibility are likely to be one of the factors facilitating long-term coexistence of both host and virus in nature. However little is known about the mechanisms and dynamics of resistance in environmental host–virus systems. In this special issue Tomaru and colleagues (2009) report on a phenomenon of reversible resistance of a dinoflagellate to an RNA virus and suggest that an intracellular suppression mechanism is responsible.

As the cost of sequencing declines, more comparative genomics studies are being carried out and are revealing striking similarities between viruses isolated on the same host at different times and places (Angly et al., 2009; Ceyssens et al., 2009; Redder et al., 2009; Weynberg et al., 2009) indicating that, if the environment is adequately sampled, we will eventually gain an understanding of the viral diversity existing in nature. Clearly viral genomes are not static, and as mentioned above, their evolution is affected by the transfer of genes between hosts and viruses (Sullivan et al., 2009; Weynberg et al., 2009), as well as by mutation and recombination between viruses (Marston and Amrich, 2009; Redder et al., 2009).

Finally, in this special issue, Kropinski and colleagues (2009) put forward a rational scheme for the naming of newly isolated bacterial and archael viruses for us to consider and Brüssow (2009) presents an intriguing historical perspective on the animal origins of many human viruses.

Over the past decade, new and exciting findings in the field of environmental virology have spearheaded a recent revival into research of bacterial viruses. Furthermore the discovery of distinctly different archael viruses has opened the door to a whole new realm of viral research. Although the known archaeal viruses reveal an exceptional degree of diversity with regard to both morphotypes and genomes, this might still be an underestimation: they are presently the least studied component of the biosphere and have only been isolated from a limited number of habitats (nearly exclusively from geothermal and hypersaline environments).

Despite the recent increased interest in environmental viruses, our knowledge remains sparse. We are still only scratching the surface of discovery of global viral diversity, have little understanding of the functionality of the majority of genes in the global viral gene pool and the roles they play in the interaction with their hosts, and are still grappling with understanding viral impact on ecological and evolutionary processes. The decade to come promises much on these fronts.