How come direct cell-cell spread of HIV allows ongoing replication in the face of antiviral therapy?

Our struggle with HIV/AIDS epitomizes societies' millennia-old fight with microbial pathogens. One goal of HIV research is to generate effective interventions that will allow us to: 1) prevent further spread of HIV (vaccines and behavioural changes), and 2) eliminate the  virus from those already infected (antivirals).

This sterilizing immunity - as it is referred to - has been a long sought-after goal for a number of viruses, yet widespread use of highly-active anti-retroviral therapy (HAART)  fails to completely remove the virus from an individual patient. Somewhere, somehow HIV is continuously replicating in your body. But how and why is this possible?

The spread of HIV from Dendritic cells to T lymphocytes or T to T cells may allow escape from antivirals. http://pathmicro.med.sc.edu

One potential mechanism is that HIV lies in a latent state - anatomical (central nervous system) or biochemical (following DNA integration and before gene expression) - where these drugs cannot inactivate it.

But researchers, headed by David Baltimore at the California Institute of Technology, have come forward with both theoretical and experimental evidence suggesting a novel mechanism that explains how the virus may be able to circumvent HAART treatment through continuous replication and direct cell-cell spread. Their results were published in Nature last week. See here.

What is cell-cell spread?

Viruses can infect new cells via a number of mechanisms. The most well-characteristic being via cell-free virus particles (see computer simulation paper here). In this, new viruses are released from the originally infected cell and diffuse to an uninfected one nearby, thus establishing a novel infection.The problem with this is that it is pretty inefficient. For one, the viruses could be carried anywhere and even if they reach a cell, it may not be the right one. One other, more efficient means of transmission is through direct cell-cell spread, which generally takes diffusion out of the picture.

Model of how HIV moves from cell-cell. A) Dendritic cell (DC) with natural fold-like projections. B) DC picks up virus particles in red C) virus is held in vesicles within cell, D) T cell projections induced, E) formation of the virological synapse, F) release, binding and entry of HIV into T cells. (Felts, et al 2010)

HIV and its close relative, the human T lymphotropic virus (HTLV-1) have been shown to move from one cell to another through this mechanism.

Original paper here

Initial visualisation

In-depth 3D structural analysis

Check out the great videos here


Following interaction with dendritic cells or T cells, the virus directs the assembly of a structure (referred to as the virological synapse) linking the two cells via alterations of the cytoskeleton which causes the two plasma membranes to come close together. This structure initially derives from the target cell.

It is here where new virus particles are released into the gap that has formed, thereby increasing the efficiency of spread by directing transmission and limiting the effects of diffusion.This synaptic structure, ultimately acts to concentrate both virus and receptors at defined subcellular locations. This kind of spread may allow the virus to become essentially invisible to our bodies' immune defences, such as neutralizing antibodies, and even antiviral drugs.

How does this allow escape from treatment?

Their model: a) inefficient cell-free infection will be eradicated with antvirals. The more efficient direct cell-cell spread will persist as it involves many more virus particles. b) The mathematical model (backed up by experimental data) showing the loss of infection (transmission index) with increasing concentration of antiviral (TFV) when we have few (m =0.2) or many (m = 100) virus particles.

Expanding the model

Initially, they generated - and experimentally verified - a mathematical model of how HIV transmission may differ during antiretroviral treatment under cell-free (inefficient - low virus transmission) or direct cell-cell spread (efficient - high virus transmission) mechanisms. Their results indicate that when there are many viruses around, the infection is more resistant to the drugs. This, they suggest, is due to the increased probability that at least one virus particle will not interact with the antiviral.


This observation extended to direct cell-cell spread when they introduced previously infected cells with non-infected ones and measured transmission of virus. Both the experimental and theoretical modelling suggest that this spread could be responsible for the ongoing replication seen in patients being treated with antivirals.





What does this mean for HIV therapy? Well, this work brings experimental evidence on the already known process of ongoing HIV replication in the face of antivirals. Through experimentally identifying this mechanism, the researchers here have uncovered a weak-point in HIV biology, one that is not currently being targeted. This method of cell-cell spread may also facilitate escape from neutralizing antibody responses through physically preventing their interaction with HIV proteins. How then might our own bodies combat this virus if its ability to more from cell-cell was inhibited? This work further adds to the growing body of work highlighting direct cell-cell spread as a principle mode of transmission for these retroviruses.

Sigal A, Kim JT, Balazs AB, Dekel E, Mayo A, Milo R, & Baltimore D (2011). Cell-to-cell spread of HIV permits ongoing replication despite antiretroviral therapy. Nature PMID: 21849975

Viruses hitch-hike through your body along the immune cell highway

Nipah virus. 

Imagine this: Rhinoviruses - one of the culprits responsible for the common cold - enter our body through the upper respiratory tract yet here it stays; the initial and generally the only site of replication is the nasal epithelium. This is how we get a runny/stuffed nose. Contrast this with a virus like nipah virus - a deadly and re-emerging pathogen spread by bats and found across South-east Asia that also enters our body via the upper respiratory tract yet leads to infection and disease in a number of our organs, including the kidneys, blood vessels and the brain, resulting in fatal encephalitis.

How come one virus remains localised while the other goes systemic? And, more specifically how does it transport itself throughout the body?


Well, virus infection of a host is a complex multi-step process involving initial contact and entry into the organism (through the nose) , early replication in particular easy-to-access tissues (upper respiratory tract) and then in some cases the spread to specific tissue sites throughout the body (brain). It is these two stages of replication that most often or not lead to the development of disease yet just how does the virus traverse the gap between the two tissue sites - especially given the minute size of a virus particle?

There are probably three hypotheses of virus spread that could be correct here: lots and lots of virus particles could be released directly into circulation (lymph fluid and blood) from the early site of replication and our blood circulation could do the work for it; the virus could infect those immune cells that cluster around sites of virus infection - or are present naturally in the initial site; or finally, the virus could basically cling on to those highly motile immune cells and be trafficked around the body and effectively transfer infection to the blood vessels and other organs.

Could your white blood cells transport virus throughout your body?

One group has recently asked this question with reference to nipah virus (read the paper here) and has discovered that this virus doesn't infect human immune cells although it does bind to them and this virus/cell  interaction facilitates infection of other cells and may allow systemic spread.

They initially came at this problem at an in vitro level - albeit using cells taken directly from the blood of healthy volunteers. The group must have initially thought that nipah must directly infect white blood cells and this is how it spread - after all this is fairly common for other related viruses, such as measles. To determine exactly which cells supported virus replication they added a green fluorescent protein (GFP) - expressing virus to a panel of white blood cells derived from the blood of the healthy humans and specifically assayed for virus-mRNA synthesis, GFP expression and how much virus was released into the culture medium. Surprisingly only one cell type - dendritic cells (DC's) - an antigen-presenting cell - supported  any kind of replication and even then it wasn't great (see below).

GFP-nipah virus infects control neuronal U373 cells but not human immune cells - except DC's to an extent

So how come nipah isn't so good at infecting these cells? Is it a receptor issue? Well the group looked at the mRNA levels of the two nipah virus receptors (Ephrin B2 and B3) in all the cells under investigation and found little correlation between their expression and the ability of nipah to infect them. For example, even the dendritic cell which had the lowest level of both receptors is able to support entry and replication while the other cells (macrophages and monocytes) that express higher levels of it fail to do so. The authors hypothesize that the DC's are engulfing nipah virus particles via a process known as macropinocytosis instead of via virus/receptor binding.

I mentioned earlier that the virus doesn't actually need to infect the cells to use them as an effective means of transportation - it can really get by through binding to the outer membrane of the cell much like a microbial hitch-hiker. So they looked at how much nipah virus was associated with each cell following stringent washes and surprisingly all the cells looked at were able to bind nipah virus particles even when they failed to get infected themselves. 

But what exactly is blocking infection when the cells bind virus AND express receptor molecules on their surface - something is inhibiting entry. The paper doesn't really address this issue but points to a role of a virus receptor-independent molecule that binds to nipah virus particles yet prevents internalization and engulfment. And even more interestingly, these virus-laden cells were able to efficiently transfer the infectious particles to other cells - as shown with the DC's and PBL's below and this ability to 'trans-infect' was retained over a couple of days (see below).

Transfer of infection with virus bound to immune cells overlaid  on top of other cells

OK, so all this work really paints a nice picture of virus infection in the host through the interactions with certain white blood cells that stably bind to - yet fail to get infected themselves - and hence are able to transfer these infectious particles to other cells throughout the body. But this is all cell culture work - no animals have been worked on here so how are we to know if this actually occurs during infection in vivo? Well, the group performed an experiment where-by the mixed nipah with hamster white blood cells and then injected these virus/cells back in to the animals and finally observed whether disease occurred and if so, how bad was it?

Hamster infections with nipah or nipah bound immune cells.

As you can see opposite, by just re-introducing the white blood cells into the animals no death occurred while directly injecting virus into them resulted in 100% mortality but then when the virus/cell mixture was added, these cells were able to transfer the infection to the hamsters with 50% mortality following acute neurological disease reminiscent of that which follows human infection.
So these cell-bound viruses are able to transmit infection in vivo, which points to it having a role in humans.

Now with all this information to hand we can now develop a model as to how nipah infects and causes disease following spread within our bodies - this is known as pathogenesis (see below). Virus could initially enter our body through dendritic cells found within certain epithelium (as the virus could infect these cells in vitro) and then these DC's could traffick to local lymph nodes where virus particles could be released and bind to the white blood cells found there; these cells would then go about their business moving around the body thus transferring infection to a range of different cell types (endothelial cells within blood vessels for example)

Current model for nipah infection and spread within the body

Just one final thought: watch this video below first - imagine this cell going about its normal routine of moving along your blood vessels and squeezing through them but with it covered in infectious virus; every cell ut encounteers will more than likely be exposed to virus and potentially become infected. No wonder nipah is such a deadly pathogen.

Mathieu, C., Pohl, C., Szecsi, J., Trajkovic-Bodennec, S., Devergnas, S., Raoul, H., Cosset, F., Gerlier, D., Wild, T., & Horvat, B. (2011). Nipah Virus Uses Leukocytes for Efficient Dissemination within a Host Journal of Virology, 85 (15), 7863-7871 DOI: 10.1128/JVI.00549-11