Revealing Virus-Host Interplay

See allHide authors and affiliations

Science  01 Jul 2011:
Vol. 333, Issue 6038, pp. 45-46
DOI: 10.1126/science.1208557

Discoveries made by environmental virologists during the past decade or so have revolutionized our perception of the living world. It has become apparent that viruses are the most abundant living entities on Earth, outnumbering their hosts by an order of magnitude. The vast majority of viruses infect microbes, but many infect humans, making each one of us a platform for a complex microbial community. Researchers now recognize viruses as major players in small- and global-scale ecosystems (1), boosting their interest in studying virus-host interactions (2). Our inability to cultivate the vast majority (>99%) of microbes under laboratory conditions, however, has limited study of these interactions. We still do not know the hosts of most viruses. On page 58 of this issue, Tadmor et al. (3) take a considerable step toward overcoming this limitation, reporting on an approach to identifying virus hosts that does not require the culturing of viruses or host microbes. The approach, which features a genetic-analysis technology called microfluidic digital polymerase chain reaction (PCR), adds to recent developments in culture-independent, high-throughput technologies (4, 5) that promise to provide a revealing picture of dynamic host-virus relationships.

Tadmor et al. endeavored to understand the interplay between uncultured viruses and bacteria that were harvested directly from a natural environment, the hindgut of termites. After harvest, they diluted the bacterial cells and loaded them into a PCR array panel so that each chamber was either empty or contained only a single bacterium. Then, they used microfluidic digital PCR (6) to assess how a selected viral gene marker was associated with a selected microbial gene marker. As bacterial markers, they used genes encoding the small subunit rRNA, a vital component of ribosomes. Based on previous analyses of microbes living in the termite hindgut (7), they chose as the viral marker a gene that encodes the large subunit of terminase (TerL), an enzyme essential for viral particle assembly (8). TerL is found only in certain evolutionarily related “tailed” bacterial and archaeal viruses, as well as in eukaryotic herpesviruses (8, 9).


Microfluidic digital PCR techniques enable researchers to document virus-host associations without culturing viruses or host microbes. The technology could help reveal virus-host interactions by creating a bridge that links culture-dependent and culture-independent methods.


Tadmor et al. detected 41 “colocalizations” of the viral and bacterial markers. In 28 of these cases, the viral marker was associated with just four microbial phylotypes (evolutionarily related hosts); all were members of the spirochetal genus Treponema. This identified these bacteria as hosts for the viruses. Tadmor et al. also grouped the Treponema-infecting viruses into five clades based on phylogenetic analyses of TerL. Analyses of these virus-host systems revealed that certain viral clades are differentially associated with certain bacterial clades, suggesting that neither cross-species transmission of viruses nor horizontal gene transfer between viruses is frequent in these systems.

The greatest merit of the method of Tadmor et al. is that it can be employed to examine virus-host interactions in biological samples from virtually any environment, without the need to enrich or cultivate organisms. As the authors note, by itself metagenomics (the analysis of genetic material recovered directly from environmental samples) “has as yet done little to shed light on the nature of specific viral-host interactions.” However, combining metagenomics with microfluidic digital PCR will undoubtedly advance environmental virology (see the figure. Due to the viral gene marker they selected, Tadmor et al. were restricted to studying certain tailed bacterioviruses. However, selecting alternative markers could help to finally expose the hosts of other, poorly understood viruses. For example, metagenomic analyses have revealed that uncultured single-stranded DNA viruses of the family Microviridae are highly abundant in marine environments and might play an important ecological role (10); microfluidic digital PCR technology could identify their hosts.

So is it time to abandon microbial cultivation? Does the identification of uncultured virus-host systems and infection patterns take us to the frontier of understanding these interactions? Not really. Rather, the new technologies represent essential steps in this quest. The richness of virus-host interplay (1113) can be unraveled only when the system is experimentally tractable both in nature and in the laboratory, and when researchers are able to apply the full range of genetic, molecular, and structural biology techniques. High-throughput methods should not carry us away from culture-dependent research; instead, the two approaches should be merged into a continuum.

References and Notes

  1. Supported by the European Molecular Biology Organization (ALTF 347-2010 to M.K.) and the Academy of Finland Center of Excellence (grant 11296841 to D.H.B.).
View Abstract

Navigate This Article