I have written quite a bit about Cytotoxic-T-Cell-based HIV-1 vaccines here on ERV. Though antibodies can target HIV-1 viruses, and HIV-1 infected cells, CTLs should be the go-to workhorses for killing HIV-1 infected cells.

Because HIV-1 is a virus, it uses all of the host cells machinery to produce babby viruses. Because of that, host cells process and present pieces of HIV-1 proteins in their MHC Class I molecules. CTLs should cruise by, take one look at those chunks of HIV-1 proteins and say “THAT AINT RITE!” and persuade the infected cell to commit suicide.

But I just had a horrible thought while reading a paper on CTL-based HIV-1 vaccines… What if some people will never be able to train their immune systems to look for HIV-1 with a CTL-based vaccine?

Vaccine-induced HIV-specific CD8+ T cells utilize preferential HLA alleles and target specific regions of HIV-1

This paper isnt about what Im going to talk about at all– But the data in this paper is what got me thinking about this topic.

They looked at the MHC Class I alleles of people who had gotten a CTL-based HIV vaccine– we all have different kinds of MHC I. I might present Chunk A of HIV, you might have a different kind of MHC I and present Chunk B, and someone else might present Chunk C. It might turn out that presenting Chunk B means you can control HIV-1 infection really well, compared to Chunk A and Chunk B people.

But apparently, some MHC I alleles ignore HIV-1 proteins. They dont present Chunk Anything. There is no chance in hell of ‘training’ those individuals CTLs. I mean, certainly an aspect of vaccines is that some people have more/less protective response to the same vaccine than others… but this is extreme. What if there is a percentage of the population that would never make a protective response from your vaccine, while others might benefit greatly from it?

You have more than one flavor of MHC I– but about 45% of us worldwide have some kind of an allele A*02. Only 22% of the A*02 people who received the CTL vaccine presented a chunk of HIV-1 in that MHC. B*07, only 11% presented HIV chunks, B*08 0%. Zero. Other flavors of MHC I presented just super– 100%. Others, ZERO.


We are so screwed on so many levels.


  1. #1 Kevin
    July 7, 2011

    I doubt anyone has an answer to this, but what levels of successful vaccination would we need before heard immunity kicks in? I know squat about epidemiology, but it seems like it would be far lower for an STD (especially now that it can be treated) and something like flu.

  2. #2 Amenhotepstein
    July 7, 2011

    Some great intelligent design ya got there!

  3. #3 Amenhotepstein
    July 7, 2011

    Uh, I meant the immune system there, not the experimental design. You see, it was kind of a joke about intelligent design creationists and, see, they always think all biological systems are perfect and, uh….

    My commenting skilz is teh best!

  4. #4 ERV
    July 7, 2011

    Kevin– Its different for every vaccine-virus combination.

    It depends on the efficacy of the vaccine, how many shots you need (and how much immunity you get from an incomplete cycle), how the virus is spread/seasonality/etc. It also depends on where you live– population density, weather, if the virus was endemic in your area, etc.

    If you just want a rough estimate, 70% is on the very, very low end of the scale, 90% would be more realistic.

    Amen– I know rite! If we are made in Gods image, then God is a retrovirus, and our immune system appears to be perfectly designed… for a retrovirus…

  5. #5 Tristan
    July 7, 2011

    This kinda relates to an idea that I’ve thought about in relation to the pathology of celiac disease, and how we might use it. Basically, celiac disease arises because a little peptide from gluten, gliadin, has two key properties – it binds very, very strongly to a couple of MHC haplotypes, and it gets covalently bonded by transglutaminase to various human proteins in the gut.

    The result is these ungodly hybrid peptide fragments which present chunks of human proteins that would normally not get presented, and so they’re recognized by the body as foreign and induce an autoimmune response. Stop eating gluten, presentation stops, response goes away.

    Anyway, this got me thinking: when you attempt to vaccinate against HIV, for the most part you get useless antibodies because most of the peptides presented are in hypermutable regions of the viral proteins (or, as you show above, there may not be anything presented. But we know that function-blocking antibodies are possible, because some people manage to make them.

    So, why don’t we do the same trick as in celiac: take the best guess sequences for the production of function blocking antibodies, and fuse them to peptides that we know will actually be presented. Downsides (if the approach works at all) would be that you’d need a whole library of such fusions, personalized to each different MHC type, and that your “vaccine” would have to be injected every few weeks – because just like in celiac, as soon as it’s cleared from the system the antibody levels drop.

  6. #6 DrDuke
    July 8, 2011

    Don’t be so pessimistic. Even a vaccine that prevents infection in 50% of people can wipe out HIV in a matter of a decade or two. The bigger problems are delivering the vaccine to people who need it most. We’ve had several polio vaccines capable of eliminating polio since the 1950s and yet we have not been able to vaccinate the entire world.

    What percentage of people produce zero CTL response to the vaccine? If it is 80% that is a big problem, but if it is less than 2% it is not so bad.

  7. #7 Geoff
    July 8, 2011

    Basic epidemiology 101: Herd immunity is dependent on the “basic reproduction number (R0)” for a virus…basically, an estimate of how many people one infected person transmits the virus to. If each infected person transmits the virus to one other person, the R0 = 1. If the R0 is greater than 1, the virus spreads in a population; if R0 is less than 1, the virus dies out. The “herd immunity threshold” (qc), or the percentage of the population needed to be immune for the pathogen not to spread in a population, is defined as 1 – (1/R0). So if the R0 = 2, qc = .5, or 50%. HIV-1 is actually a pretty shitty transmitter, with estimated R0 values between 2 and 5, suggesting that the qc is between .5 and .8 (compare this to measles, which has an R0 of ~14-18). Of course, one other way to get HIV-1 to die out is to decrease its R0, by focusing on prevention. Providing clean needles to IV drug users and condoms to commercial sex workers would also dramatically decrease the R0, at a cost much less than developing highly effective vaccines!

  8. #8 Heather
    July 8, 2011

    Very interesting post! It may be alarming to realize that some current therapies used to treat disease are suboptimal. We’d like to think that new drugs and vaccines will result in cures for all sorts of diseases, but the truth is that not everyone’s immune system functions equally.

    As a fellow PhD biology graduate student, this post is particularly interesting because it highlights the importance and amazing nature of scientific research! Clearly, this data leads us to believe that more research must be done in order to design more effective vaccines to fight HIV (although vaccines should be used in coordination with other preventative measures to increase the chances of elimination of the virus). For example, it is unclear if the MHC haplotypes, which apparently “ignore” HIV proteins, are expressed by cells that may not being able to adequately process and subsequently present the HIV PEPTIDE to CD8 T cells. In which case, it may not necessarily be that these MHC molecules “ignore” HIV proteins, but that they cannot even pick up the peptide from inside the cell to present it! Perhaps there is a way to engineer a vaccine that didn’t rely on protein processing and presentation, but consisted of purified immunogenic peptides. A CD8-based vaccine is essentially useless if it can’t activate the CD8 T cells, which will only happen when the T cell recognizes PEPTIDE (not protein) presented by MHC.

    Importantly, a variety of factors contribute to one’s ability to respond effectively to a vaccine: heritable genetics, current or past infections may also alter the ability to fight viral infections or respond adequately to a viral vaccine, and even your sleep pattern!

    I recently wrote a post regarding newly published data by Lange, et al. about how sleep deprivation prevents vaccines from working to the best of their ability (Hepatitis A vaccine was used in these human-volunteer studies). If you’re interested in discussing and learning about the data behind scientific breakthroughs regarding human health, disease, and immunology I invite you to check out my blog @ http://escapinganergy.blogspot.com/ !

  9. #9 Richard Jefferys
    July 8, 2011

    Also seems that HIV epitopes restricted by “protective” HLA-B alleles may be disappearing over time:

    Loss of HIV-1 derived CTL epitopes restricted by protective HLA-B alleles during the HIV-1 epidemic.

  10. #10 Mary
    July 11, 2011

    HIV does always seem to be 1 or 2 steps ahead of us….

    *puts on cheerleader outfit* HOWEVER..

    we have made inroads….yeah!

    …I lived in Ptown in the early 80’s…watched men dying on the street…most of the men in my graduating class are dead…

    *takes off cheerleader outfit and sets fire to it…**

  11. #11 Drew
    July 12, 2011

    I remember a paper that came out in 2004 or 2005, I think, that said that HIV (IIRC it’s specifically the Nef gene product) actively cycles MHC-I molecules back off the cell surface (specifically HLA-A, B, and I think C but not E which also means that the cell won’t be the victim of NK killing). Doesn’t that suggest, then, that relying on CTL is not likely to work unless you’re combating a Nef deficient virus?

    Incidentally I always had this idea of trying to attenuate and use a Nef deficient virus as a live-attenuated vaccine. I wasn’t the only one that thought about this but trials were denied…apparently regulators had a problem with it, some crazy notion about giving a “live” HIV-virus to patients being a bad thing. But it worked in macaques and SIV (again IIRC, it’s been a while since I looked into it and my recollections could be WAAAAAAAYYYYYYYYY off, especially given that wasn’t, and still isn’t, my specific area of research and was simply reading it as an interested “outsider”).

  12. #12 Drew
    July 12, 2011

    Actually I was mixing up some different papers. First was Cohen et al. 1999 Immunity 10:661-671 This one showed that HIV-1 down-regulates HLA-A, and B, but not C or E.

    Next was Williams et al. 2002 J. Virol 76: 12173–12184 which showed that Nef bound the cytoplasmic tail of HLA-A2 but not E and channeled it away from the surface.

    Finally was Kasper et al. 2005 J. Biol Chem 280:12840-8 suggested that the disruption actually happened in the trans-golgi compartment and shunted HLA containing vesicles to the endolysosomal pathway.

    Now I’m not suggesting that CTL immunity is not a necessary part of any particular attempt at an HIV vaccine, in fact I think that it would be pretty important given that infected cells also escape antibody detection by also reducing surface presentation of HIV gene products. It all just gives me pause, thinking that with all the immune evasion that HIV does I just don’t know that there will ever be an effective vaccine…I still support the effort to make one.

    Incidentally, as an undergrad I also had this crazy idea that if you could catch HIV infections early on, and give a treatment that would wipe out the immune cells, specifically CD4+ T-cells, and monocyte/macrophages effectively cutting off the main centers for replication (Yes I know that there are additional locations for replication) the remaining viral replication could be dealt with by conventional therapies (anti-retrovirals and INF-a). You’d of course need to figure out some way to reconstitute the immune system, no small feat in itself.

New comments have been disabled.