Pure Pedantry

ResearchBlogging.orgIt’s amazing what the kids are up to these days.

This one comes live from Mount Sinai (my present educational residence). Hubner et al., publishing in Science, use an infectious, fluorescent strain of HIV to watch the virus move from one cell to another. Their results are fascinating and may help us develop better ways to treat the disease.

(Full disclaimer: This research was performed in the Chen lab at Mount Sinai where my roommate presently works.)

It’s funny how my biases work. I mean, I am not a microbiologist, but here is the bias that I had about how infections like HIV work: I figure that you have large quantities of virus floating around in your blood. These viruses invade whatever cells they happen upon, forcing them to produce more virus. Eventually those cells explode (lysis) from too much virus, setting them free into the blood to wreak more chaos. We could call this the carpet-bombing theory of viral action: limited selectivity, maximum damage.

It turns out that HIV doesn’t work like this (mostly). In fact, it operates more much more sneakily — like special forces — viral ninjas, if you will. Instead of spreading out in the blood, HIV viruses transfer between infected cells through a structure called a virological synapse. (To be accurate, HIV does infect cells in a cell-free form — this is discussed in the Introduction of the paper. However, cell-to-cell transfer of HIV is up to a thousand times more efficient and inhibiting it inhibits viral replication.)

Virological synpases are assembled like synapses between neurons. They are formed from the binding of specific proteins on the surface of the joined cells. These proteins travel and are organized on the surface of the cell using the cytoskeleton. In the case of virological synapses during HIV infection, these synapses organize the viral receptor proteins on the CD4 and CXCR4 as well as the viral envelope protein Env. When HIV hijacks a cell’s activity to make more virus, it triggers the production and assembly of all the molecules required to make a new virus at this synapse. Then all that is required is to link with another cell.

Here is where the crazy part comes in. Hubner et al. use a special fluorescently-labeled form of HIV to watch the virus assemble at the virological synapse and transfer from one cell to a human T-cell.

How, you ask? First, the Chen lab created the first — to our knowledge — fluorescent, infectious strain of HIV. It is fluorescent because one of the structural proteins of HIV, Gag, was modified to include a fluorescent protein called GFP. GFP fluoresces green and can be viewed under a special microscope. I say infectious because while many strains of fluorescent virus have been created, none have been capable of assembly such that you can watch them through all phases of their life cycle.

After they created this special HIV, they took cells that they had infected with their virus — called Jurkat cells — and mixed them with human T-cells. Then they filmed the cells using a very expensive microscope that takes pictures every couple of minutes for several days. When you put all those pictures together, you can watch two cells come together, form a synapse, and then watch the HIV go from one cell to another.

An example of one of those videos is below:

The white cell is the donor cell, a cell already infected with HIV. The round bright spot is the virological synapse between cells. Not labeled is T-cell cell attached to this cell. You can see how little puncta, “buttons,” of white float away from the synapse into the other cell. This is HIV being taken from one cell to another.

Here is another video:

This is a 3D reconstruction. What the microscope they use does is take optical slices through the cells — 2D. These can be reconstructed to form a 3D image. The video is such a reconstruction being rotated. The donor cell is green. The round green splotches are virological synapses. The red cells are the recipient cells.

So aside from the videos being super-cool, what do they tell us about how HIV works? Well, in addition to the technical advance, this paper had several interesting findings from watching HIV go from one cell to another. Here are two:

  • First, there is the actual form that HIV gets into a recipient cell. If you remember from the videos, HIV going into the new cell got there in puncta or “buttons.” Rather than just entering the new cell and filling it entirely, at the beginning the HIV huddles together. The authors attempted to determine the composition of these bundles. Partly using electron microscopy, the authors found that these bundles represent vesicular compartments in the recipient cell. The recipient cell is enveloping the virus with its membrane and pulling those vesicles into itself — a process called endocytosis. These vesicles are remarkably stable. Presumably, the virus eventually escapes these vesicles and infects the cell. This is intriguing because it means that it we could prevent that process of envelopment, we may be able to fight the HIV. Further, it is interesting because a variety of other viruses — an example is influenza — also use this strategy to infect cells.
  • Second, the authors explore why infection by HIV prefers this cell-to-cell route. A little background: patients with HIV can develop antibodies that prevent cell-free transfer of the virus. They may be not be able to fight off the virus completely, but they can prevent this route of transmission. If you think about it, this is not entirely surprising. If the immune system encounters free virus just floating around, it isn’t that hard to make an antibody for it. In contrast, HIV spends most of its life cycle inside T-cells. There it is much more difficult to identify and to stop. Interestingly, the authors watch infection with HIV both in the presence and absence of patient serum containing antibodies that block cell-free transmission. They found that these antibodies did not stop cell-to-cell transmission. By going through cell-to-cell transmission, HIV evades one means the immune system uses to stop it.

In the grand scheme of things, this paper represents an important — albeit, incremental — advance in our understanding of HIV. Scientists are working to understand every step of HIV’s life cycle and trying to find drugs to inhibit every one. If indeed much of viral transfer happens at synapses between cells, then drugs created to stop this step need to take this into account. I have no idea how you would create a drug to stop synapse formation or endocytosis into the recipient cell, but I am sure that is what people are starting to think about. Every time we discover a new step, scientists trying to cure the disease are one step closer.

Congratulations to the Chen lab on this very interesting paper!

Hubner, W., McNerney, G., Chen, P., Dale, B., Gordon, R., Chuang, F., Li, X., Asmuth, D., Huser, T., & Chen, B. (2009). Quantitative 3D Video Microscopy of HIV Transfer Across T Cell Virological Synapses Science, 323 (5922), 1743-1747 DOI: 10.1126/science.1167525

(Incidentally, the paper had many, many cool videos that I couldn’t help not putting up. Below are two especially cool ones.)

This shows a cell-to-cell transfer. Left is just imaging the HIV. Right is showing where the cells are plus HIV in green. Note how in the surrounding T-cells slowly get the HIV puncta.

This one shows an actual cell being infected. See how it turns green. It’s important that it comes into contact with an infected cell, and then becomes infected.


  1. #1 Abhishek
    March 30, 2009

    Simply awesome!Apart from the obvious significance in terms of furthering research into HIV, I think these images have the potential to become iconic – could be used widely in films and documentaries, much like the short clips of cell division and artificial insemination that come up in movies remotely related to the topic of cloning.

  2. #2 Max
    March 30, 2009

    Can’t help but wonder what the denialist response is going to be.

    I had read a Scientific American article some years ago that discussed how HIV reproduces massively upon infection and the immune system responds immediately but slowly gets eaten away. I couldn’t envision a model for how that worked. This helps immensely.

  3. #3 Chris Noble
    March 30, 2009

    Can’t help but wonder what the denialist response is going to be.



    Henry Bauer said

    Martin, Michael, mo79uk:

    That video whihc has aroused such a fuss shows, first, only that a T-cell has been injected with a large amount of dyed protein. After some time, a clump of that accumulates and then breaks off from the cell, and appears to attach itself to the outside of another cell. It isn’t shown entering that cell, let alone showing the clump breaking up, which is what infection would involve: stripping the protein coat, releasing the RNA, then RNA –> DNA in the cell’s nucleus.

    It’s not even clear what that clump consists of besides dyed protein. Maybe it’s one of those microvesicles filled with foreign protein that the cell just wants to get rid of.

    One of Chen’s (co-author of the film work) responses to Michael cites, as a good explanation for lay people, the story at http://whyfiles.org/shorties/280HIVinfects/
    The text there says that for HIV to spread in the body, each dying T-cell must infect more than one other T-cell. On the other hand, the electron micrograph shows several “HIV buds” ALL AT THE “SYNAPSE” WITH JUST ONE OTHER CELL


    resistanceisfruitful: If you thought Photoshop was awesome, check out this video of cellular messaging being depicted as HIV “clone” infection.

    I have some thoughts about this, but would prefer others more knowledgeable than me do so first.

    followed by

    whereisthetruth: I guess that because they have to inject a radioactive gag to make it fluorescent

  4. #4 alan silver
    March 31, 2009

    Got to your blog via NYT saw that you had a Ghostbusters quote…as a former Mt Sinai faculty member now engaged in quality-of-care surveillance and improvement (among other things)…a favorite of mine

    [Dana is possessed]
    Dr. Peter Venkman: I make it a rule never to get involved with possessed people.
    [Dana starts passionately making out with him]
    Dr. Peter Venkman: Actually, it’s more of a guideline than a rule…

  5. #5 Roberta
    March 31, 2009

    What is fascinating from this is whether viruses do spread via synapses–and whether neurons borrowed this or were created by viruses, the way mitochondria were from bacteria. Can you ask your roommate about this, I am really interested.

  6. #6 Barry
    April 1, 2009

    I wonder if the pope would be interested in seeing these videos.

  7. #7 rijkswaanvijand
    April 2, 2009

    Synaps describes form more than function in this case.. As a structure for telecommunications between cells neural synapses evidently only serve purpose in multicellular organisms. Neural communication is basically selective iontransport within and between highly specialized cells. Ionchannels seem to have become an integral part of single-celled organisms pretty early in evolution, and subsequently came to play key roles in many processes, including formerly mentioned.

  8. #8 Penny
    September 15, 2009

    This is a really nice blog! I’ve been reading an AIDS textbook (published in the 90’s) that describes the possibility of cell-to-cell transmission of HIV and I couldn’t find any current information online that was as easy to understand but informative as this!

New comments have been temporarily disabled. Please check back soon.