Viruses ! Is there any word related to microbiology that conjures up the same fear, the same cache of woe than viruses. The very definition can mean "a harmful or corrupting influence", which is presumably why Miley Cyrus's music is often transmitted via "Viral Videos". People know there is something to fear when they hear of a Virus.
So perhaps it is not mistake that so close to Halloween that the Microbiology Twitter Journal Club is getting together to talk about a new species of GIANT VIRUS.
So how does one go about hunting for huge viruses?
The Giant viruses that we know of tend to infect micro-organism known as Acanthamoeba. This soil dwelling Amoeba is one of the most common types of protozoa found in the soil, and survive by eating the bacteria that live within water and soil sediments. Giant viruses survive by preying upon them.
So let us get back to our hunt for these viruses. The researchers took their soil samples and then mixed them into a special buffer which contained a cocktail of Amphoteracin B* and antibiotics (Penicillin and Streptomycin) to kill off all fungi and bacteria in their samples.
They added these sediment samples to cultures filled with Acanthamoeba. If there were viruses present that preyed on Acanthamoeba, they would start killing them off.
They found that sediment samples from the Tunquen River in Chile could kill off Acanthamoeba in this manner.
They also took samples of freshwater from an unidentified pond outside of Melbourne, and used a thick solution of polyethylene glycol, which somehow can dissolve viruses very well**. The purified virus samples were then used to infect Acanthamoebae, and found to kill them off.
When the Acanthamoebae exposed to their samples died off, theoretically releasing multitudes of viruses into their medium, the researchers further purified the samples to use to re-infect new cultures of Acanthamoeba.
This brings us to our first set of figures. First they took pure cultures of the viruses, and then imaged them using a microscope. This is what they look like:
The Virus in A1 was the one recovered from Chile, named as Pandoravirus salinus, and the virus in A2 was recovered from Melbourne and named as Pandoravirus dulcis. This confirms that the viruses are indeed huge. Most viruses we know of are far too small to be seen using a light microscope, so the fact that these particles are even visible is extraordinary.
So they then looked at Acanthamoeba under a light microscope to see whether these tiny viruses could infect them.
Both viruses could be seen within the Acanthamoeba.
So the next step here is to take a look at the infection cycle of these particles, to prove once and for all that these are indeed viruses, and not some esoteric form of bacteria.
They needed to get far more detailed pictures of the viruses for this, so they used a powerful electron microscope to look at the virus infection cycle.
Step 1. The Virus gets itself eaten by the host.
Acanthamoeba naturally consumes bacteria within its environment, engulfing them in structures known as vacuoles, where they can be killed and be broken down into nutrients. The virus enters its victim via a similar route, but once it ends up in a vacuole, it declines to be digested. Instead, it fuses its surface with the membrane, and injects its DNA into the host cell's cytoplasm. This fusion is being pointed to by an arrow.
Let's zoom in to this fusion, so we can get a better view of what's going on.
The dark area A is a dense mesh of fibrils that coat the virus. It is less clear what structures relate to areas B & C. But what is clear is that there is stuff being transferred from the virus membrane to the host membrane.
Step 2: Eclipse Phase
When the virus is nothing more than a shell and it's contents are swimming in the host cytoplasm, we can no longer the actions of the virus directly. This is often referred to as the eclipse phase.
But during this phase, we do observe changes to the cell, in particular we can see the nucleus begin to behave in a bizarre way.
In Figure S1 A we can see what a healthy nucleus looks like in the centre of the cell.
In Figure S1B, the virus has infected the cell, and is causing the nucleus structure to transform. The Nucleus becomes less dense, and its membrane develops folds, and a whole new structure forms near it (bottom left corner) which has an unknown function.
As the infection progresses, this structure and the remnants of the nucleus disappears, and we begin to see viruses form in the cell's cytoplasm (Figure S1 C).
If we really zoom in on these new structures we can see a strange lattice structure forming within them, although I'm not enough of a microscopy expert to tell whether these are real structures or digital artefacts.
Step 3: New Virus Production
The Researchers took a closer look at the viruses that were now being constructed within the cytoplasm of the host. The construction appears to start from a distinct point in the virus, and the internal components are pulled together at the same time as the external capsule proteins. The virus appears to be knitted together in the cytoplasm.
The host membrane then rapidly fills with these Pandoravirus particles, until it eventually bursts.
The researchers have shown the infection life cycle of these viruses using electron microscopy. But that is only the first part of the paper. The next stage involves working out how these viruses are related to all of the other forms of life we currently know of.
Origins of the Pandoravirus.
The researchers first needed to establish the degree to which these viruses were related to eachother. So they extracted all of the proteins from each virus, and ran them on a gel to get an idea of the different types of proteins on a gel based on their weight.
From this, we can see that both viruses contain protein, but that they have proteins of different weights, and they both produce one protein at around 62 kDa in large amounts. So we can probably say that P. salinus and P. dulcis are quite different viruses. We'll come back to these proteins later.
The researchers delved deeper, by extracting out DNA from presumably pure preparations of Pandoravirus. They used three different sequencing machines, each of which work in slightly different ways. They used an Illumina HiSeq 2000 machine, a PacBio RS and the soon to be extinct Roche 454 GS FLX+ machine.
They found that the main difference between P.salinus and P.dulcis related to their genome sizes, with P. salinus having a much larger genome. This is mainly due to the presence of four genome chunks (highlighted in green) that are present in P. salinus but not P. dulcis.
These chunks on further analysis were found to consist mostly of gene replications.
Because these viruses are so similar, the researchers decided just to focus only on the Pandoravirus with the most genes, P. salinus. With more genes to work with, it would potentially provide a better basis for analysing the relatedness of this parasite to other species.
This is where thinks get very difficult. This is organism is very different to any that is already known, and this is reflected in its genome.
Most of the genes have no viable match.
This is where we go back and take a closer look at the proteins extracted from the virus in Figure 1C. They used mass spectroscopy to identify the proteins from the virus, specifically ones from the coloured portions of the pie chart above. This confirmed that the virus does indeed produce proteins in the same way as its host organism, but deepens the mystery over the huge amount of proteins it encodes that we have never seen before.
The researchers noted that this virus lacked key genes needed for cell function, which is a characteristic for most viruses. Specifically , they note that the virus can solve the problem of these "missing" genes by co-opting much of the machinery found within the cell nucleus, providing a neat explanation of why the nucleus of the Acanthamoeba disappears during infection. Backing this up was further studies of the proteins, and finding that the purported transcription machinery of the virus isn't included in the particle, suggesting that the host nucleus plays a large role in infection here.
The researchers do some good work studying the replication cycle of the virus, and uncovering the secrets of its genome. The big story here is the sheer amount of what is unknown about this virus. Approximately 93% of all the genes found in the Pandoravirus cannot be traced to any cellular lineage.
The researchers suggest that this could be evidence that these viruses evolved from an ancient bacterial ancestor that is entirely extinct today, and that perhaps this could indicate that there was a fourth domain of life that became extinct. Whilst I really like that explanation***, I'd need to see a lot more data.
I don't really have many criticisms for this paper. The only question I would have is to do with how they did their transcriptomics. I get the impression that they extracted RNA from infected Acanthamoebae cells, but have no idea how they would tell the difference between the normal RNA produced by Acanthamoeba and the viral RNA. It could be that they had specialist software to compare these things, or some other experimental technique to isolate the viral transcriptome.
On the other hand, I did find some of their speculation on genome/protein function interesting. At some points in the paper, they speculate on the nature of this virus based on the genes which they couldn't find in the genome. This would be all well and good if approximately 93% of the genes in the genome were of an unknown origin, and therefore unidentifiable.
I don't want to sound crazy here, but what if the functions of the genes missing from the 7% of the genome were performed by the 93% of the unknown pat of the genome ?
But then again, it is fairly understandable that the researchers would make this speculation. It mostly agrees with what we know of the data, it just takes that extra step.
This paper something new and interesting and focussed solely on genomics and microscopy, which means I don't have to make any comments at all about statistics this week. I liked it.
EDIT 2/12/14: I got an E-mail from a reader who had some issues with this papers methods. As I suspected, this paper was not free from flaws, but the flaws were simply beyond my ken.
If you'll read the paper carefully once again, I'm sure you'll notice the authors didn't do any "transcriptome analysis" at all (no RNA extracted either). As far as I can see, they only worked with DNA and proteins extracted from purified virus particles. Moreover, the only time they mention "transcriptome" explicitly, this is for desirable experiments to be done in the future: "A comprehensivetranscriptome analysis will be requiredto identify all the intron-containing genes ..." When they discussed some introns before, that is, this was solely based on 'in silico' analyses, done from DNA in comparison with the few genes showing definite homology to other organisms and letting the software extract the information about the tentative Pandora introns.
*Note: Amphoteracin B is also a treatment for Protozoal infections, and the researchers had to acclimatise their strains of Acanthamoebae through what I refer to as the "Dread Pirate Roberts" method, in which they gave small increasing doses of the poison to the amoebae until they developed a resistance.
** I couldn't find a study to explain exactly how it works. There have been suggestions that the shape of the virus has an effect, but I could get a clear answer from the literature.
***I wouldn't even call it a "fourth domain", I'd call it an "Nth" domain because I'm sure that there were many different domains out there before our last universal common ancestor came onto the scene.
Philippe N., Legendre M., Doutre G., Couté Y., Poirot O., Lescot M., Arslan D., Seltzer V., Bertaux L. & Bruley C. & Pandoraviruses: amoeba viruses with genomes up to 2.5 Mb reaching that of parasitic eukaryotes., Science (New York, N.Y.), PMID: 23869018