Instructions for a retrovirus vaccine? Not really

When folks in my lab think about biological problems, we think about basics. A pathogen has some molecular component that trips a sensor on the outside of a cell. That sensor (the receptor) grabs on to some adapter proteins and starts a cascade of chemical reactions catalyzed by various enzymes, which eventually leads to the activation of a transcription factor or factors that changes the behavior of a cell. We think at the level of the cell, at the level of the protein and at the level of the molecule. Occasionally, if we write a paper and the reviewers insist, we might do an assay to see if the effect that we're seeing changes the ability of a cell to activate T-cells.

But when I need to tell someone without a background in biology why we research what we do, my go-to answer is vaccines. I say that if we understand the way that our cells recognize a pathogen and initiate the immune response, then we will be able to make vaccines that are more powerful, targeted and effective. In truth, there's no clear line (at least to me) between my work and the clinic, but I believe we need basic research to lay the groundwork for ground-breaking practical discoveries.

This paper takes the opposite approach:

ResearchBlogging.orgInnate Immune Sensing of Retroviral Infection via Toll-like Receptor 7 Occurs upon Viral Entry

Rather than carefully dissecting the signaling pathways, they do an end-run around the biochemsitry I usually care about and go right to the functional output. I mentioned that we occasionally (after a lot of arm-twisting), look at T-cells. For these guys, their starting point is one step beyond T-cells: antibodies.

The story of this paper is fairly simple. Most inbred mouse strains used in labs are susceptible to infection with two retroviruses: Murine Leukemia Virus (MLV) and Mouse Mammary Tumor Virus (MMTV). But one mouse strain, called I/LnJ, is not. These mice make a robust immune response that results in neutralizing antibodies and clearance of the viral infection. Since we've mostly failed at making effective vaccines against retroviruses (the most notorious of which is HIV), these guys wanted to know which cellular sensor (receptor) was initially detecting these viruses. Many different receptors are capable or recognizing viruses, but if there's one that's particularly important, it would be good to know so we can target vaccines to that receptor directly. Indeed, they found that if you knockout Toll-like receptor 7 (TLR7) or the adapter protein that it uses to signal (called MyD88), the antibody response in these mice is completely eliminated.

But for everyone hoping for a solution to HIV vaccination, there's not really cause for celebration just yet. This is a single mouse strain and only two retroviruses. Sure, we have to start somewhere, but there's no reason to think that it will be the same for other retroviruses and other mice. And speaking of other mice, these authors did not actually address why their mouse strain (I/LnJ) is able to mount an immune response and the others aren't. Other mice have TLR7 too, but they get sick and die when infected. It does us no good to know how this mouse strain detects retroviruses if detection alone does not normally lead to an effective immune response.

Furthermore, I'm not convinced that TLR7 is the only receptor that's important. The argument in this paper is that there are only two toll-like receptors that recognize RNA (retroviruses have RNA genomes). They knocked them both out, and TLR3 wasn't required, but TLR7 was: case closed. The trouble is, toll-like receptors are not the only types of receptors that can see RNA. There are also RIG-I like receptors (RLR's - I apologize for the acronym salad, but there's really no other way to do this). The author's don't consider RLR's because RLR's reside on the inside of cells. The authors believe that the recognition occurs before the virus gets inside cells, but don't actually provide any data to support that claim. TLR7 is certainly important, but many immune responses require input from multiple receptors, and the authors don't rule this out.

Finally, just because TLR7 is the natural method of recognition, this does not mean it's the only method of recognition that could be used for a vaccine. In fact, the authors have an argument against this idea in their own data! At one point, they use complete Fruend's adjuvent, a commonly used immune-activator (that doesn't rely on TLR7), and this also leads to an antibody response. So all this paper really tells us is that I/LnJ mice are special when it comes to retroviruses (which was known), and TLR7 is necessary for this speacialness (which is new), but we still don't really know why.


PS - I almost forgot - one more thing on the HIV front: these guys are looking at the production of antibodies, which makes sense because most of the effective vaccines we have these days are based on antibodies. But if you read ERV, you know that antibody vaccines, even if we figure out how to make them, may not be effective against HIV anyway.

Kane, M., Case, L., Wang, C., Yurkovetskiy, L., Dikiy, S., & Golovkina, T. (2011). Innate Immune Sensing of Retroviral Infection via Toll-like Receptor 7 Occurs upon Viral Entry Immunity, 35 (1), 135-145 DOI: 10.1016/j.immuni.2011.05.011

More like this

You've got "the argument in this paper is that" twice in a sentence in the "Furthermore" paragraph. [Fixed! thanks :-) -KB]

So what are the good things about this study? If the receptor they looked at doesn't seem to have anything to do with resistance directly, why is knowing it's involved useful?

Well, the good thing about this study is that they identify a receptor that's necessary for detecting these retroviruses and mounting an immune response. It will probably be cited a bunch, and a lot of people can probably use these conclusions as a jumping off point for interesting biology about the immune response to retroviruses.

This is how science works, incremental advance, standing on the shoulders of giants etc etc. The problem I have with this paper is that I think they took good, informative results and then stretched their conclusions farther than was warranted by the data. This is what it takes to get into a good journal these days (and it worked!), but still.

So, the way they determined TLR7 was essential was by knocking it out and seeing that the immune response didn't happen, right? How is that helpful if you're trying to go the other way and stimulate the response? Or was that the whole point of your criticism?
Could you explain (generally) the hypothetical chain between receptor identification and vaccine?

@ Rachel - It can be helpful because we have synthetic products that we know act through TLR7. So if

1) retroviruses normally signal through TLR7 AND
2) activation of TLR7 is the best way to get an anti-retroviral response AND
3) TLR7 is the only receptor necessary to get that optimal response

THEN we could potentially use those synthetic chemicals as adjuvents in vaccines to trick the immune system into making a good anti-retroviral immune response.

I think this paper showed that (1) is true for this mouse strain and these retroviruses, but not much else.