Fake virus particles generated by fluorescence microscopy Patrick Forterre 1, 2 , Nicolas Soler 1* , Mart Krupovic 2 , Evelyne Marguet 1 , and Hans-W. Ackermann 3 1 Institut de Ge ´ ne ´ tique et Microbiologie, Universite ´ Paris-Sud, CNRS UMR 8621, 91405 Orsay Cedex, France 2 Institut Pasteur, Department of Microbiology, 75724, Paris Cedex 15, France 3 De ´ partement de Microbiologie-Immunologie and Infectiologie, Faculte ´ de Me ´ decine, Universite ´ Laval, Quebec, QC, G1K 7P4, Canada Many laboratories are actively studying the abundance and roles of viruses in natural ecosystems. In these studies, the presence and number of viral particles is usually determined using fluorescent dyes. However, DNA associated with membrane-derived vesicles (MVs), gene transfer agents (GTAs), or cell debris can produce fluorescent dots that can be confused with viral particles. We suspect that fluorescence counting can lead to overestimation of virus numbers and even sug- gest the presence of viruses when there are none. Future studies in environmental virology should acknowledge this point and consider how to bypass this problem. Besides trying to improve discrimination between vir- ions and MVs, we suggest adopting less holistic approaches, focusing on the detection of known virus groups and the isolation of new viruses from a broader range of hosts. Finding viruses in the environment In recent years, the interest in viruses as an integral part of the biosphere has been growing and several authors have proposed that viruses have played a much more active role in shaping biological evolution than it was previously believed [1–5]. In particular, ecological studies have sug- gested that viral particles (or more precisely virus-like particles, see below) are the most abundant biological entities on our planet [2,6,7]. Many laboratories have been actively studying the roles of viruses in regulating popula- tion dynamics and geochemical cycles in a variety of eco- systems, including humans. Viral ecology is thus becoming a thriving new branch of microbial ecology, itself an expanding discipline that attempts to combine traditional (cultivation) and molecular approaches such as shotgun metagenomics. The number of viral particles in environmental studies is usually assessed by fluorescent staining, with SYBR Green being the most common dye used to label viral nucleic acids [8,9]. Fluorescent dots that are supposed to represent viral particles are visualized by epifluorescence microscopy (EFM) or counted by flow cytometry. Metagenomic analyses focusing on viromes (see Glossary) also rely on EFM as the method of choice to validate the presence of viruses and the absence of cellular contamination in environmental samples before DNA extraction and sequencing [10,11]. EFM, which is fast, easy, and inexpensive, has completely replaced the classical approaches of virus enumeration by plaque assay or by transmission electron microscopy (TEM) [12–14]. Theoretically, the use of EFM provides the possibility to count viruses that cannot be cultivated in laboratory con- ditions, and to discriminate between virions containing nucleic acids and nucleic acid-free virus-like particles. Con- taminating free DNA is usually removed by DNase treat- ment [11]. The use of EFM was indeed a breakthrough in viral ecology and generated numerous publications, often in high profile journals [15]. However, the exclusive use of fluorescence staining in counting viral particles also raises several problems. A well-recognized one is that EFM is not efficient in detecting RNA or single-stranded DNA viruses [16]. This is an important issue because virome analyses have shown that these viruses are much more abundant in natural environments than previously expected [17,18]. We would like to focus here on another, less frequently Opinion Glossary Extracellular DNA: DNA present in the environment. This DNA can be attached to biological or mineral particles and can account for a large proportion (up to 70%) of all DNA present in a given environment. Gene transfer agents (GTAs): particles resembling virions of tailed phages capable of encapsidation and transport of fragments of cellular DNA. Kill the winner hypothesis: predicts that increase in the abundance of a specific host triggers significant increases in viral infection and subsequent lysis of this particular species, thereby controlling its abundance. Membrane vesicles (MVs): globular vesicles (50–200 nm in diameter) produced by budding of the outer membrane in proteobacteria or the cytoplasmic membrane in archaea. They can fuse with recipient cells and carry cellular or plasmid DNA. Membrane vesicles have morphologies very similar to those of some spherical tailless viruses. Microfluidic technology: enables control and manipulation of the behavior of fluids in microstructures. This technology has been successfully adapted for various single-cell studies. Virus-like particle: general term that a priori covers true virions, GTAs, and MVs. Unfortunately, it is also often used as a synonym to virion or viral particle. Virocells: virus-infected cells actively producing virions. Unlike regular cells, virocells cannot further divide. The virocell concept posits that infected cells are the organismal form of the virus. Viromes: the complete set of viral genomes present in a given environment. Corresponding author: Forterre, P. (patrick.forterre@igmors.u-psud.fr) Keywords: epifluorescence microscopy; viral ecology; virome; membrane vesicle; gene transfer agent. * Current address: Institut Cochin, INSERM U1016, CNRS UMR 8104, Universite ´ Paris Descartes, 27 rue du Faubourg Saint Jacques, 75014 Paris, France. 0966-842X/$ – see front matter ß 2012 Elsevier Ltd. All rights reserved. doi:http://dx.doi.org/10.1016/j.tim.2012.10.005 Trends in Microbiology, January 2013, Vol. 21, No. 1 1