Dengue versus the epigenome

Dengue virus (DENV) particles. Does dengue interact with the host epigenome? And, why?

Epigenetic modification of chromosomal structure has the ability to rapidly and stably alter gene expression and function within cells, tissues and whole organisms. And, changes have been found to induce a number of diseases in humans, such as cancer and others.

The ability of infecting viruses to influence this process has now been realized with a number of large DNA viruses being shown to take advantage of host epigenetic modifications, including: the histone proteins that are necessary for the structure and modulation of chromatin.

But what hasn't been seen is whether RNA viruses, which replicate within the cytoplasm are able to at least interact with the host chromosome structures. And, more importantly, why would they do so? Researchers have just published evidence suggesting that dengue virus interacts directly with host histone proteins and requires such activity for replication (see PLoS paper here). The reasons why are not understood.

DENV-C specific histone binding (binds to H2A, B, 3 and 4) but not GFP. Does this has an affect on host function?

Dengue Virus (DENV) is a small, single-stranded, positive-sensed RNA virus and belongs within the Flavivirus genus, alongside yellow fever virus and west Nile fever virus. It is spread between humans via the bite of the female Aedes mosquito and can lead to the development of very high fever (41 degrees Celsius), headache and in some cases, a lethal hemorrhagic fever.

It is estimated that two-fifths of the worlds' population are at risk of DENV infection and it is endemic in over 100 countries within tropical or sub-tropical climates. There is currently no clinically proven treatment for dengue fever nor are there any vaccines available, making any research into the basic biology of the virus all the more important.

Using a 'tagged' protein, this group showed that a recombinant DENV C protein (capsid protein forming the outer layers of the virus particle - yellow in the above picture) interacts with the four core histones, H2A, H2B, H3 and H4 found within the nucleus of huh-7 cells, a human hepatocellular carcinoma cell line used a lot in DENV research. See the figure above for the specific binding. DENV is known to infect human liver cells, especially in fatal cases.

GFP-tagged DENV-C co-localizes with huh-7 nuclear and cytoplasmic histones (H2A and H2B)

When over-expressed within these cells, a GFP-tagged C localised to the same areas of the nucleus as each of the histones did and this was extended to DENV-infected cells (see above). C also interacted with the histones while in the cytoplasm as well. In vitro C and histones formed dimers together in the absence of nucleic acids and it was subsequently shown to bind DNA with - or without - histone proteins. See below.
 
This interaction was shown to disrupt the normal oligomerization of histones but not that of histone-DNA binding. Infection up-regulated the expression of the core histones and was essential for normal DENV replication.

DENV-C binds to huh-7 H2A, B, 3 and 4 and causes formation of oligomers (the bands higher up on the gel - compared to lower and smaller bands).

The physical basis of this interaction is currently not understood but is believed to result from the structural similarities between the histones and DENV-C. A common folding pattern - important for oligermization - is shared by both and the authors state they are in the process of crystallizing the two together. The biological significance of this observation was not really addressed here, apart from the fact that histones were required for DENV replication (see graph below). The authors hypothesize the reasons as to why DENV might need this process, but provide no evidence to back it up.

DENV replication requires H2A and H3 proteins. Shown here by siRNA knockowns of the two (columns 1 - 4) when compared to non-specific controls (5 and 6).

It would be interesting to see if this interaction applied to other human cells, including primary cell lines and maybe in vivo. And, if it extended to mosquito cells as well? But, where in the genome does DENV-C bind? Is it specific or more general? What is the basis of the requirement for these histone proteins? Finally, is it possible to inhibit this relationship to develop new anti-DENV therapies.


Colpitts, T., Barthel, S., Wang, P., & Fikrig, E. (2011). Dengue Virus Capsid Protein Binds Core Histones and Inhibits Nucleosome Formation in Human Liver Cells PLoS ONE, 6 (9) DOI: 10.1371/journal.pone.0024365

Should vaccinology embrace systems biology?

Any intelligent fool can make things bigger, more complex, and more violent. It takes a touch of genius -- and a lot of courage -- to move in the opposite direction.

E. F. Schumacher (1911-1977) British Economist

Vaccines represent one of the most cost effective methods around to prevent loss of life and disease in a whole range of animals, including the human population. Over the last 200 years or so, we've become pretty adept at producing them and so the science of vaccinology - or how to generate these complex pharmaceuticals  - has led to the eradication (and near eradication) of many viral pathogens.

This is one network in your immune response following influenza vaccination (Kokke et al 2011) as id'd through systems biology approaches. Knowledge of key mediators in these pathways may allow for the rational design of new vaccines - but is it worth it?

But, it hasn't succeeded for a number of currently killer viruses (respiratory synycitial virus and HIV to name but two) and we have begun to think that maybe the method of 'isolate, attenuate, vaccinate' or the synthesis of single virus antigen molecules isn't gonna cut it anymore. So what are we going to do?

We are wanting to rationally generate vaccines - taking a wild-type, disease-causing isolate and through some genetic engineering, make it sufficiently weak so as to generate an effective immune response in patients while not causing disease. Yet, this is harder than it looks and so some researchers are now turning to systems biology to offer a glimpse into how some of our most successful vaccines function so as we can reconstruct these processes for the numbers of viral pathogens we are yet to protect against.

*I have explored how we may attempt to do this from the viruses perspective (mumps virus versus vaccine, here), more precisely: how come the vaccine strain is less deadly than the 'wild' strains. Both are valid approaches and probably just as difficult to carry out as each other.*

Systems biology affords us a chance to more fully understand the complexity of living systems. Through the collection of reams of data (DNA sequences, gene expression changes, protein levels - and other 'omics' technologies) we are now able to adequately model what is going on in the organism/cell through now more common bioinformatic and statistical analyses. As in the quote I used above, it is not about making the study of life more complex, it is really about realizing this fact and doing something to understand it better under a broader way of thinking. This allows us to 'see' changes and functional differences that we would never have observed had we gone about such an experiment using our a priori knowledge and this global, holistic view may just be the savior that vaccinology needs now.

Example of the complex data collected during a typical 'systems' experiment - what does it all mean, and how can we find something important and worthwhile to study?

I am aware of a number of papers that are currently using this process as a primer to develop improved vaccine products (see here, here and review here of virus vaccine examples). These guys - for example: Bali Pulandran of the Emory Vaccine Centre in Atlanta, USA - are interested in comparing the immune response (humoral response, innate immunity and gene expression changes) of human subjects administered with vaccines. The response to the yellow fever virus vaccine as well as two types of influenza vaccines have been approached and through complex bioinformatic modelling they were able to pull out some significant correlates of immune response - this they hope will aid in the future testing of novel vaccines and facilitate a rational take on vaccine generation by identifying a gene(s)/protein with a functional role in the immune response. This they have begun to do in some of ther papers above - it is nice to see this kind of work being used as a basis for experimental biology.

These types of studies hail a new way of thinking about viruses, vaccines and the immune response to them; if only we can realize the power in taking a step back and looking at the diversity in each. And this type of work could be applied to any number of mechanisms such as vaccine safety or applying it to different tissues during an infection.

Saying that - this stuff isn't particularly easy, cheap or quick as you might think. But as each study generates so much data, might it not take but a few such investigations to lead us on the way of rationally attenuated and protective vaccines? So, should vaccinology embrace systems biology? I think if you have the abilities to do such a study - which from a pharma perspective definitely yes, do it as the more information we have at our disposal the better position we are in. We await further results from these groups to compare how well systems thinking goes up against human ingenuity, that has worked well in the past.

Nakaya HI, Wrammert J, Lee EK, Racioppi L, Marie-Kunze S, Haining WN, Means AR, Kasturi SP, Khan N, Li GM, McCausland M, Kanchan V, Kokko KE, Li S, Elbein R, Mehta AK, Aderem A, Subbarao K, Ahmed R, & Pulendran B (2011). Systems biology of vaccination for seasonal influenza in humans. Nature immunology, 12 (8), 786-95 PMID: 21743478

Infectious disease meets the epigenome

The study of epigenetics (see Pharyngula's excellent article) has allowed us to see biology and genetics through new eyes. The fact that heritable traits can be encoded in not only the nucleic acid sequences of As, Cs, Ts and Gs but also in the physical conformation of chromosomes and chemical modification of DNA has added a new level of complexity to our understanding of life. Covalent modifications to both DNA and associated histones and chromatin can result in the formation of active or repressed genetic regions; transcription of these genes found in that area is thus activated or repressed. Embryonic development, behaviour and cancer formation  have all been impacted by the discovery of this new genetic system wit deregulated epigentic processes leading to the development of these diseases - but what about in infection, immunity and pathogenesis of associated diseases?

[caption id="" align="aligncenter" width="300" caption="Epigenetics: DNA, histones, ovalent modification and chromatin."][/caption]