A Giant Among Giants
by Merry Youle
Without a doubt, Mimivirus is remarkable. For a virus, it is extraordinarily large and complex. But it is hardly one of a kind. The more that researchers look for large viruses, the more they find.
Although phages generally tend to have small genomes, some managing with but a handful of genes, a glance at the current NCBI list reveals that there are now eight with sequenced genomes that amount to more than 200 kb. A Pseudomonas phage tops the list with 317 kb, but the not-yet-sequenced genome of Bacteriophage G of Bacillus megaterium is reported to be ~670 kb.
Figure 1. A colorized SEM showing numerous particles of the phycodnavirus PBCV‑1 attached to a Chlorella NC64A host. Bar = 500 nm. Source
Serwer and colleagues have pointed out that the procedures used to isolate phages are biased against the giants. Typical plaque assay protocols call for at least 0.3% agarose in the overlay. However large phages, such as Bacteriophage G, can't make plaques of significant size when the agarose concentration is 0.2% or higher. Using 0.15% agarose, these researchers isolated a novel Bacillus thuringiensis phage from soil — the first representative of a new group of large myoviruses. It too has a genome with more than 200 kb packaged inside a 95 nm capsid that sports a tail measuring half a micrometer!
Figure 2. EM of novel Bacillus thuringiensis phage from soil — the first representative of a new group of large myoviruses. Bar = 0.1 μm. Source
Most of the known giant viruses infect eukaryotes and are members of a monophyletic group known as the nucleocytoplasmic large DNA viruses or NCLDVs. They earned the "nucleocytoplasmic" label because they either replicate entirely in the cytoplasm or initiate the process in the nucleus and then complete it in the cytoplasm — thus independently of the host's transcriptional apparatus. Here you find the pox viruses of vertebrates and invertebrates, the phycodnaviruses (phyco- meaning algae) of marine and freshwater algae, and the amoeba-infecting Mimivirus. Phycodnaviruses of note include the coccolithovirus that plays a role in the termination of blooms of an abundant marine alga, the coccolithophore E. huxleyi, as well as a large virus that infects Ostreococcus tauri, the smallest known free-living photosynthetic eukaryote.
Figure 3. (Click to enlarge) Phycodnavirus OtV5, with a 122 nm capsid, infects Ostreococcus tauri, ~1 µm in diameter. TEM. A At high multiplicity of infection (moi), many viruses can adsorb to a single cell. B Virus replication results in the accumulation of viral particles in the cytoplasm before cell lysis occurs. Bar = 500nm. Arrows = virus particles; Chl = chloroplast; Cyt = cytoplasm, n = nucleus, m = mitochondrion, sg = starch grain. Source
The phycodnaviruses are quite remarkable themselves. Their genome lengths are mostly in the 300 kb range, but one is 560 kb. The archetypal phycodnavirus that infects chlorella-like algae has ~373 protein coding genes — more than the number often touted as the "minimum" required to support life. However, gene numbers don't tell the whole story as these viral genomes lack many of those listed in the "essential" gene set. In an unvirus-like manner, this genome also encodes 11 tRNAs and three kinds of introns plus genes for multiple DNA methyltransferases and DNA site-specific endonucleases — the enzymes that make up the restriction modification systems found in many Bacteria. These genes are functional; all chlorella viruses have methylated bases in their genomes, each virus with its own characteristic site-specificity. And most intriguing of all, this chlorella virus has the gene needed to synthesize hyaluron, and synthesize it it does, eventually covering its chlorella host with a dense fibrous network. Hyaluron synthesis had been thought to be an art exclusive to vertebrates (and a few pathogenic bacteria that include it in their capsules to fool our immune system). Even more bizarre, some chlorella viruses make chitin instead, and yet others make both and accumulate both on the surface of their host.
Figure 4. (Click to enlarge) Infection of Chlorella NC64A by PBCV‑1. B Attachment of PBCV‑1 to the algal wall and initial digestion of the wall. D Complete digestion of the algal wall. F An empty viral capsid remaining on the surface of the host. Bar = 100 nm. Source
Mimivirus has the phycodnaviruses beat by just about any yardstick you choose, and it even crosses that illusory line intended to separate viruses from cellular life. Its ~500 nm capsid is larger than the smaller bacterial cells such as Mycoplasma. Its 1.2 Mb genome contains 981 predicted protein-coding genes — double the number found in the smallest known Bacteria (Mycoplasma genitalium) and Archaea (Nanoarchaeon equitans). But a virus it is, firmly placed phylogenetically within the NCDLV group, albeit on its own branch.
We don't know what most of those 981 genes do as they lack identifiable homologs in the sequence databases, but at least 95% of them are transcribed during infection. Where did they all come from? Many appear to be paralogs produced by gene duplication events in Mimivirus. Of those with clear homologs, most are related to bacterial genes, a few to genes in Acanthamoeba and other protists. These likely came via horizontal gene transfer from a host, from other parasites present in the host, or from Bacteria phagocytized by the host for food.
Figure 5.Left: Mimivirus infecting an amoeba. The virion has been phagocytosed and resides within a vacuole. The inner membrane of the virion (light circle) will later fuse with the vacuole membrane to discharge the virion contents into the cytoplasm. Right: close-up view. Credit: Didier Raoult. Source
Having a 1.2 Mb genome presents some challenges. One is simply synthesizing enough DNA for >300 progeny viruses in about 12 hours. In one experiment, researchers measured a 7‑fold increase in total DNA within the host in the first 8 hours, so recycling of host DNA by viral endonucleases simply won't suffice. Not surprisingly, Mimivirus (and other NCLDVs) encodes numerous enzymes for nucleotide metabolism and synthesis. Next comes the task of packaging those genomes into the preassembled capsids, a process that takes place in a cytoplasmic "virus factory." The unique "stargate" that opens upon infection to rapidly deliver the genome is a story in itself (click here and here).
The protein capsid measures ~0.5 µm; the dense layer of reticulated polysaccharide fibers covering it surface increases the diameter to ~0.75 µm. It was the faint Gram-positive staining of those fibers combined with the virion size that earned Mimivirus its name: Microbe mimicking virus. Such large size may be necessary to efficiently infect amoebae and other protists via their feeding phagocytosis pathway. Studies using precisely-sized beads found that individual beads greater than about 0.6 µm are taken up immediately, whereas smaller ones accumulate on the cell surface until combined they reach sufficient volume to trigger uptake.
Figure 6. A model of the complex virion of Mimivirus (cross section viewed perpendicular to the unique five-fold axis). From the outside in: head proteins (black) and shafts (green) of the surface fibers that are attached to the anchor proteins (blue spheres) that cover the lattice forming the icosahedral capsid (red spheres). Next, an additional protein/lipid layer (gold), unidentified fibers (orange), and the bilayer lipid membrane (green). Inside the membrane is the genomic DNA (black) with associated proteins (green) and other proteins (pink). Thick blue lines on the surface represent the stargate. Source
With such fascinating stories being told by Mimivirus and the other giants, people are now looking for them in more environments. Modified techniques are called for, as those used previously to spot viruses may have excluded many of them. For example, when collecting marine samples for viral metagenomes, researchers often use filters with 0.16–0.2 µm pores to catch the "microbial" fraction and allow the "viral" fraction to pass through. Realizing that many NCLDVs are apt to be caught with the microbes, Monier and colleagues searched the "microbial" sequences from the Sorcerer II Global Ocean Sampling (GOS) Expedition for NCLDVs using their conserved DNA polymerase sequences as a handle. They found Mimivirus sequences in 86% of the samples and chlorella viruses in a third.
Claverie and colleagues see no limitations that would preclude the existence of even larger viruses. Unlike cellular organisms, there are no metabolism-based constraints on particle volume. Of course, a virus must be smaller than its host, and Mimivirus is <1/30 of the size of its host amoeba. Bacteriophage G may be approaching the limits here: a 200 nm diameter phage infecting a 2 µm Bacillus. Given that Mimivirus can fit 1.2 Mb of DNA into its 0.5 µm diameter capsid, they surmise that a virus with a 10 Mb genome would be possible. It would require only a 1 µm capsid, a size easily accommodated by large amoeboid protists.
What do you think is the likelihood that Mimivirus will still be #1 giant five years from now? We'd bet not, as much of the virosphere is yet to be explored, and likely there is more than one researcher with dreams of discovering the next viral leviathan. Large protists that feed on bacteria would be the place to look.
References
Van Etten JL. (2003). Unusual life style of giant chlorella viruses. Annual Review of Genetics, 37, 153−195. PMID 14616059
Claverie JM, Ogata H, Audic S, Abergel C, Suhre K, Fournier PE. (2006). Mimivirus and the emerging concept of "giant" virus. Virus research, 117 (1), 133−144. PMID 16469402
Amazing stuff! You realize that, sooner or later, we are going to find viruses that contain some kind of innate metabolic activity within—say some kind of ATP generating ability necessary to keep its genome packaged properly. Then the paradigm will not simply crack a bit!
I also think a great deal about the intentional selection strategies we microbiologists use. The percentage of agar in top agar described here is a great example.
I wonder what else we are assuming in our work at the bench?
Notice again, friends and colleagues, we don't see this kind of information in textbooks! It's going to take many sources of information to teach the wonders of the microbial world, including blogs like this one.