Can we visualise virus infection as it happens - in real-time?

One worthwhile way to study viruses – and other micro-organisms – is to see where exactly they are found within a host. How do they enter the body? What organs do they infect and how? How do they spread from tissue to tissue and organ to organ? How do they exit the body? These are just some of the questions which it would be good to actually SEE how and where it happens. Maybe then we could better understand the dynamic relationships governing infection and disease; maybe then we could design better, more effective therapies. Just maybe. Like it is all that easy.

How could we track adenovirus movements?

For one thing viruses are pretty small - so, are there any high resolution methods of looking at infection that may be able to help us? Pathologists can look at tissue samples from living or dead patients (animals included) and microscopically assess these for signs of viruses using for example: antibodies specific for particular proteins or nucleic acid probes (in situ hybridisation or PCR perhaps) found only in infected cells. This stuff is pretty good and is routinely used for both, clinical sciences and biological research but its kind of limited and in the last decade or so, the use of ‘reporter genes’ - like GFP and luciferase has facilitated the easier and faster analysis of viral infection.

Reporter genes are inserted into the viral genome and when expressed upon cell infection, they produce a protein – maybe a fluorescent or luminescent protein – whose function can be assayed and followed. This is what we can look at during infection. There are however certain limitations with this, in that, only those viruses that successfully infect a cell will express the reporter gene; viruses which fail to infect and are taken up by the liver or immune cells will not be detected. We are seeing a biased image of viral infection if we only consider reporter gene expression. This is particularly important when we are using viruses as therapeutic agents themselves as strict pharmacological testing requires intimate details of in vivo distribution and kinetics. So how can we see viruses in vivo without reporter genes?

As a paper in PLoS ONE, from a group at the Mayo Clinic in Rochester, demonstrates, there IS an alternative way to track viruses in vivo – the molecular attachment of individual reporter molecules to the virus itself thus requiring no gene expression and therefore allowing a more unbiased view of infection. An interesting aspect of this work is that it is carried out in real-time; these reporter molecules can be visualised as infection happens, at the millisecond scale. This also allows for the tracking at very early time points, times where no viral gene expression is taking place.

This study came at it from the angle of developing safer and more effective anticancer viruses, viruses which will infect and kill only those malignant cancer cells but this could be applied to any area of investigation. Being able to follow virus distribution in a mouse-model is of course a great scientific and clinical benefit to them. The group dyed an adenovirus vector with a molecule which emits light in the near-infrared range (particularly suited to in vivo imaging) which they then injected into groups of mice via their jugular vein. They were thus able to analyse and quantify the tissue distribution of their labelled vector throughout a whole mouse, in real-time.

This is the first time that this kind of imaging has been carried out and it certainly won’t be the last. This work could be applied to yet more viral vectors; it could be used to study a ‘natural’ infection or it could be used for non-viral imaging of therapeutic nanoparticles. A combination of this early time-point analysis with later, reporter gene expression imaging would be able to give us an unprecedented view into dynamic real-time viral infections in a number of model systems.

Brandenburg, B., & Zhuang, X. (2007). Virus trafficking – learning from single-virus tracking Nature Reviews Microbiology, 5 (3), 197-208 DOI: 10.1038/nrmicro1615

Hofherr, S., Adams, K., Chen, C., May, S., Weaver, E., & Barry, M. (2011). Real-Time Dynamic Imaging of Virus Distribution In Vivo PLoS ONE, 6 (2) DOI: 10.1371/journal.pone.0017076

Studying viral infection at the whole-organism level

Some questions about how viruses cause disease in their hosts (viral pathogenesis) are best asked and studied using an in vivo model system; just sometimes infecting cells under tissue culture conditions just doesn't cut it. Questions like: how does a virus interact with all the immune cells during an infection and what cells does the virus actually infect should be asked this way.

But of course, this in vivo stuff is a great deal more difficult than in vitro studies and appropriate animal models don't just grow on trees; this is why, when a relevant model system of viral infection comes along we get excited - well at least I get excited. Unless you look at everything in its entirety, you never know what you will miss and viruses being as small as they are, its easy to miss something important and missing something important is bad news in the world of science.



A recently published study has looked at viral infection at the 'global' or whole-organism level using transgenic zebrafish larva infected with Infectious Hematopoietic Necrosis Virus (IHNV), an RNA virus related to rabies virus and is particularly deadly if you happen to some form of salmonid. Zebrafish are generally pretty good models for a whole lot of biological processes: zebrafish genetics are pretty well understood allowing for easy transgenics; they are particulary easy to study, especially to image as they are small and transparant and some genes/pathways are well conserved with humans meaning that it may have some applications to us. These factors all suggest that zebrafish may be a pretty decent model to understand viral infection in general, not just in fish.

Following infection, they were able to look at entire whole organisms for viral presence, concentrating on what particular cells/organs contain viral mRNA  and proteins. They were able to follow infection through its entirety, at early stages and the later stages when serious disease takes hold, allowing the elucidation of intra-host viral spread and dissemination. They used their system to shed light on the mechanisms of IHNV pathogenesis, showing that viral infection led to vascular endothelium destruction and impaired blood flow. It is just near impossible or at least a lot of hard work to do this kind of analysis in any other model system.

[caption id="attachment_161" align="aligncenter" width="300" caption="Zebrafish viral infection: In blue are cell nuclei, green endothelial cells and red viral proteins."][/caption]

Using this model - as in all model systems - comes with certain caveats attached: IHNV is not a natural pathogen of zebrafish (indeed, to date no viruses have been described) , i.e. what we see here may not be exactly what happens out there in the real world when this virus infects salmon. The virus was also injected into the bloodstream of these fish which is highly unlikely to occur in the wild - how would the infection change if it were administered another way? Not considering these issues, this work offers up a decent picture of systemic dissemination of IHNV in a not-so-perfectly matched host. Only time will tell how applicable to the real-world this is.

Its hard to imagine this work being carried out in any other kind of vertebrate - transparent rats in the future perhaps? But this stuff has been carried out using GFP expressing viruses within a non-human primate model only a week before this. Although not as easy to image, hard-work and dedicatedly searching through cells and tissues for signs of infection allows us to understand viral infection at the whole-organism level more appropriate to human disease.

This pretty much makes their statement below a bit incorrect, or at least out-dated:

We describe in this paper the spread of a viral infection throughout an entire organism, something that, to our knowledge, has not been done before in a vertebrate.

As a final thought, wouldn't it be great to image viral infection in real-time using a GFP-expressing IHNV in this zebrafish model? - just checked, there is a GFP IHNV virus out there. Check out the live-cell imaging of GFP neutrophils in a zebrafish above.

Two Zebrafish larvae

Ludwig, M., Palha, N., Torhy, C., Briolat, V., Colucci-Guyon, E., Brémont, M., Herbomel, P., Boudinot, P., & Levraud, J. (2011). Whole-Body Analysis of a Viral Infection: Vascular Endothelium is a Primary Target of Infectious Hematopoietic Necrosis Virus in Zebrafish Larvae PLoS Pathogens, 7 (2) DOI: 10.1371/journal.ppat.1001269

Ludwig, M., Palha, N., Torhy, C., Briolat, V., Colucci-Guyon, E., Brémont, M., Herbomel, P., Boudinot, P., & Levraud, J. (2011). Whole-Body Analysis of a Viral Infection: Vascular Endothelium is a Primary Target of Infectious Hematopoietic Necrosis Virus in Zebrafish Larvae PLoS Pathogens, 7 (2) DOI: 10.1371/journal.ppat.1001269