Rolf Zinkernagel (2014) - Why Do We Not Have a Vaccine Against HIV or Tuberculosis?

Ladies and gentlemen it is a pleasure to be here. And to talk to you about why we do not have certain vaccines. Now there is no doubt that vaccines have been probably the most efficacious medical activity that we have succeeded to do so far. And therefore it is of course a natural thing to think if we have a vaccine against pneumonia or measles, we must also be able to make a vaccine against HIV or TB. Now I will try today to explain to you why this equation doesn’t hold. And I do so by questioning certain text book dogmas, because I had to teach immunology for 30 years at a medical school. And while reading these text books I thought, you know, at least half of what we read in text books is wrong. Of course we don’t know which half. So the first question I raise: Is memory, immunological memory responsible for protection? And the answer as you may guess is of course 'no'. And the second is, you know: What is antigenic specificity and why do those agents that we have no vaccine against, all vary basically? And of course that tells you something about the antibody specificity that is needed. And these agents simply escape an otherwise efficient immune response. So the more general conclusion of course is: Is what we measure always what we SHOULD measure? Because usually we measure with a method that gives us the answer we want to see. This may be slightly overstated but I will illustrate some of these aspects. Let’s just set the stage. You know there is no doubt in a co-evolutionary context that we only have to live for 20 or 25 years at most. Because that’s what we need to set up the next generation. And anyway, our biggest problem today on this world is of course that there are too many humans. Now the reason why we have problems in acting rationally is that our human behaviour is very inadequate. In fact I think humans are very basically stupid. And of course we have difficulties to accept that not all humans are equal. But you know otherwise there would not be any co-evolutionary selection. And of course responsibility always collides with freedom of action. And we have enough examples, particularly in application of vaccines that today one of the major problems is that certain population parts actually resist applying vaccines which in a way you know, how stupid can we get? So I think we have to work as a scientific community in educating people in very general terms. With very simple newspapers, to educate and teach them, you know what science is all about. And that actually science, for example vaccines help. Now if you look at the basic problem. And I give you now a very short survey of a minimal sort of immunology. And then try to illustrate what I want to say about specificity and about protection. Now when you look at infections with viruses - but you can extrapolate that to bacteria and classical parasites Some viruses replicate, destroy the host cell, others don’t. In fact many more are noncytopathic than cytopathic. And HIV basically belongs to that category. And you see immediately where the problem lies. If you have an immune response that is very efficient in eliminating or stopping viral replication on this side via antibody, the hematogenic spread, then such a T cell immune response, particularly by CD8 T cells, of course is actually harmful in this context. Because the virus would not kill your cell, it’s the immune response. And this basically we call immunopathology. HIV is a good example, Hepatitis B, Hepatitis C and so on are other examples. Where the immune response simply is causing the problem. Now in physiology these types of viruses jump from one host to the next at a time where the T cell response doesn’t exist. And when is that? Of course either in utero or at birth. Because we are all born without a functioning immune system. And there if the virus carrying mother hands over the virus to the offspring, there is no immunopathology because the recipient has no immunocompetence. So that’s the normal physiology. All these viruses jump at birth point. Now the second very general point I’m going to make here is that if you look at intact viral particles that actually only the tips of these glycoproteins are exposed to the surface and can be addressed by antibodies or B cells. There is simply no room for antibodies to squeeze in between these glycoproteins. And therefore simply the idea of broadly cross-neutralising or cross-protective antibodies simply is an illusion. Because these antibodies can’t get to cover up the intact viral particle, because there is no access for them. Now if you look at, this is my second immunologist slide, at the location of infections and the immunological consequences. Of course normally you think of a virus hitting your mucosa or skin then spreading to the draining lymph nodes. That’s where the first initiation of an immuno-response happens. And this usually has to be quick enough not to have overwhelming hematogenic spread of the virus. Because if the virus gets to your brain you’re dead. And that’s why early IgM responses on day 3 or 4 are so crucial. Now there are 2 alternative extremes. One is papillomavirus, warts virus. They cause a proliferation of epithelial cells. They’re out of the reach of immune cells including Langerhans cells. And therefore this event out there, basically being a benign tumour at the start, is out of the reach of any immune reactivity. Because the draining lymph node doesn’t realise that there is something going on out there. The other extreme situation is where the mother has transfused virus to the offspring via the obligatory blood transfusion at birth. And of course in this case the virus spreads all over, there is no immunocompetence. And therefore the host becomes a viremic patient but has no immunopathology because there is no immune response. And any subsequently developing immune response is simply negatively selected in so-called deletion tolerance. So that’s all for the starter. And we go to evaluate the situation on the vaccine side. If you look at successful vaccines, you know the list is here. They are all mediating their protection via neutralising antibodies. There is not a single exception to that statement. All the ones we don’t have, TB, leprosy, HIV, malaria, etc. There we need both effector T cells and neutralising antibodies. But the left hand side, your right hand side candidate all vary. This is particularly true for HIV or malaria. TB hides away, I get to that. And of course leprosy can be put together with TB. And in a way you could argue your leprosy or your TB granuloma is basically the best vaccine against these types of infections. And the chronic infection that is not killing you in terms of a TB granuloma, actually increases or so-called innate resistance mechanisms to a considerable extent. So there is no doubt that innate resistance - and we have heard, you know from several speakers about it Without interferon Alpha for example you’re dead before the virus looks at you. There is a very interesting observation that chicken eggs, but all bird eggs, are full of maternal antibodies. Which is interesting. The same is true of course for humans and mice, mothers transfuse their immunological experience in the form of antibody to the offspring. Interesting also that autoimmunity, particularly those dependent on autoantibodies, only starts about after 22 and is predominant in females versus males. And this is at least an indication that to produce lots of antibody at high titers may have a cost, and this is the reflection in autoimmune’s susceptibility. Then also tumours of course come up in general terms, solid peripheral tumours come up after 30. So that puts all the problems we cannot solve immunologically easily into the later phase. Because what kills you are childhood infections. And up till 300 years ago our life expectancy would be in the order of 25 to 28 years, for Romans it was 18 years. So that shows you that basically what kills you in an evolutionarily important way is before puberty. So let’s look at specificity. Most of immunology in text books deals with hapten’s so-called small phenolic groups like NP or DNP, dinitrophenol groups. These are very small because they basically contain 6 carbon atoms. But the interaction of a neutralising antibody with the tip of the glycoprotein of course is a much larger interaction, one speculates about 10 to 15 amino acids. And therefore it’s no surprise that the affinity of hapten-specific antibodies is in the order of 10^-5. Whereas affinity of neutralising antibodies is in the order of 10^-10. So 5 orders of the magnitude difference. And since most of immunology deals with these low affinity type of responses it’s not a surprise that basic immunology says the limiting factor to any B cell response is T help. But this is not true because the frequency of these neutralising antibody producing B cells is 10^-6. Whereas the frequency of hapten-specific B cells is 10^-2. So for all biologically relevant protective B cell responses it’s the low frequency of the B cell that is the limiting factor. Now when you look at an acute cytopathic, acutely killing virus. And I just use vesicular stomatitis virus which is a rabies-like virus and in mice behaves like rabies, is neurotropic. You see virus replicates. You have a T cell response. You have an immediate ELISA response that is parallel to the neutralising protective antibody response. For noncytopathic infections, you could take hepatitis B or HIV, you see the discrepancy that you have a T cell response. You have an ELISA response. But your neutralising antibiotic response takes between 40 and 200 days. And the same is true in a model infection in mice with lymphocytic choriomeningitis virus where you have the virus, you have the T cell response, you have the ELISA antibody response - remember at very low avidities and affinities. And the neutralising response takes 60 to 300 days. And what happens in these long slowly developing antibody responses is that while the affinity of the antibodies mature or increase, the virus by mutational escape simply runs ahead of any antibody response. And that is one of the major reasons why we don’t have a vaccine. And this in a way has a correlation to our B cell repertoire. If you take a normal serum, just from an SPF mouse or a normal human being that has not ever been exposed to a particular virus, you find usually a background neutralising titer of 1 in 20 to 1 in 40. Is this really specific or is it just sort of, you know, natural antibody? It is specific because you can cross absorb that titer by distinct stereotypically defined viruses of the same family. Now if you take a virus that does not kill you in 10 days but only kills you in 20 years, then you find there is no measurable natural or spontaneous antibody titer. Which indicates that these viruses start from a completely different repertoire usage. And of course it depends both on the virus and the repertoire whether, where and how quickly they start. So natural antibody represent or reflect the antibody repertoire and this can be illustrated. And these are experiments done by Adrian Ochsenbein in the lab. If you take mice that do not have antibodies, muMT knockouts, and you infect them with this rabies virus, they all die with a very low dose of virus. If however you give back to these mice just half a millilitre of normal mouse serum from an antibody competent mouse, you basically can correct that. And this is a very robust effect because you need now 10^8 or 10^7 viral particle to kill the mouse. So it's 4 logs difference. So natural antibody. You have them. Then you can deal with acute infections. It starts out with 1 in 30 and your protective level is 1 in 900. So it’s basically just increasing your titer by a 30-fold factor. Which is not very much but translates very easily in about the replication cycle of 6 hours of B cells once they get started. And of course the viruses that are noncytopathic can hide wherever they like. In neurons, for example with the herpes viruses, or in epithelial cells in the kidney for CMV. Or for LCMV in lung epithelial cells. And the same for CMV. Therefore there can be peripheral reservoirs that are not clinically apparent and that may well keep the immune response up like the TB granuloma and like the persisting HIV. I summarise at this point: B cells make antibodies. There is a local IgA - I didn’t talk about it but this is very important. On the mucosa there is a very primordial, ancient IgA system that actually is active in these muMT mice. Although they don’t have any serum antibody but they still have that local IgA in the mucosa. We need neutralising or opsonising IgM very early to prevent spread to the brain. The IgG depends on T help, I didn’t show that but that is very clear, it's 100% dependent on T help. And of course the IgG has a much longer half-life, IgM is in serum for about 12 hours, IgG for 20 days half-life. And it’s only the IgG that can be handed down to the offspring via the FC receptor. I don’t get into that. And basically, very interesting, there is almost no negative selection against antibodies. But that’s a different issue. Now for T cells they control and eliminate intracellular parasites of all sorts. Including and particularly important in solid organs. But to get a sufficiently activated T cell response to clear out all virus from your lung or your kidney, you would have to steam up the immune response so tremendously that your fingers probably would fall off because of an interleukin storm. So these things are well balanced and if we try to overdo them we obviously have to pay for that. The T cells are important for the long-lived IgG, but the T cells also cause immunopathology. Now why immunological memory? And is it responsible for protection?. Of course if you survive the first infection there is no need for memory because the system has shown efficient. If you die of the first infection then of course you don’t need memory either. And that is basically the starting point for discussing this issue. You see in text books 'memory' is defined as quicker and higher. So you prime, you come back with the same antigen: It's steeper, quicker and a higher titer. Now immunity as we define it is 'protection against'. And we have these positive, very efficient vaccines and we don’t have these vaccines. So what is the explanation? Well let’s look at a very simple experiment. Hans-Peter Raue did that as a PhD study in the lab. You vaccinate mice with this rabies-like virus. And then you take the so-called memory T+B cells and adoptively transfuse them into a naïve recipient. And then you challenge that recipient, either a few days later or 2 months later, they all die. So whatever memory T cells and memory B cells means, they can’t do the job. However, if you take the serum of this vaccinated mouse 3 months after vaccination and adoptively transfuse that to a recipient: Challenge, they all survive. Now that experiment of course is the experiment we all have survived at birth. Because we have received the immune IgG from our mother. And that covers basically us against all epidemiologically important infections during our first 1 or 2 years. And the mechanisms are actually quite fascinating and reflect this interesting co-evolution in humans or mice. You get FC receptors on the hemochorial side of the placenta. On the foetal side you have FC receptors and they pick up the IgG from the mother. In calves and all ruminants the situation is a bit more complicated but as illustrating, in that the placenta is a double membrane separation between the foetus and the mother. There are no double membrane transport systems for proteins. And therefore these have evolved FC receptors on their gut epithelial cells. And these FC receptors are expressed during the first 24 hours of life after birth. After that the renewal of the epithelial cells is such that the renewal cells don’t express the FC gamma receptor. So that’s why colostrum milk is so important to be given to these offspring because they absorb from the colostrums milk which basically is a dream for Nestlé, you know high concentration of immunoglobulins of both IgG and IgA type. And this actually is then absorbed by the gut epithelial FC. And that replenishes the foetal calf serum. Now that is the reason why we use foetal calf serum in tissue culture. Because foetal calf serum has no immunoglobulin. Because it cannot be transported from the mother to the offspring. And the immune, even the competence and anti-body production of the foetus of the calf is zero. So protection is via neutralising antibody. But of course if the agent is variable by mutation, you have a problem. Increase of memory B cell frequency is antigen-dependent and antigen-driven. But usually these frequencies drop back and end up you know by being 4 or 8 times higher than in an unprimed situation. But the differentiation of a so-called memory B cell to an antibody-producing B cell takes again 3 to 4 days. So even having memory B cells needs re-stimulation which is antigen-dependent. And it’s only the antigen that drives to maturation of an antibody-producing cell. And this re-exposure or antigen-driven or -dependent neutralising antibody titers can come from within. This of course is the case with persisting viruses. And measles is also noted here because measles of course cannot be isolated from immune measles survivors. But you know the disease of subacute sclerosing panencephalitis, which is a central nervous complication of wild type measles infection, and when you look there in these lesions you always find crippled measles virus. The matrix protein is usually mutated. And therefore, SSP is the clinical case but you know down the line to no clinical symptoms, you find decreasing amounts of this crippled persisting in a way vaccine. So it can come from within via antigen antibody complexes on follicular dendritic cells for 1 to 4 years. And then of course from the outside polio, all the diarrhoea viruses and so on are classical examples of that. Now let’s go to the pregnant mother. Let’s say she is genetically AB, the father is CD. So the embryo is AC and the mother could react against the C which she doesn’t effectively because there is no HLA exposure on the surface of the foetus. The mother had to have survived the virus X infections otherwise she probably couldn’t have become pregnant. Because otherwise such an infection would kill her and of course the foetus. Therefore she has anti-X antibodies. Now after birth the offspring will have anti-X antibodies and will immediately be exposed to the epidemiologically you know active infections X. And this eventually will lead to an active immunisation because the maternal antibody slowly will go down because of the half-life of 20 days. And at some point the infection will be very attenuated but still active and you will have active immunisation. If this exposure is delayed for more than 2 or 3 years then of course the first exposure will be a disaster because your maternal acquired protection is gone. And the polio epidemics in the early ‘50’s of course illustrate that very nicely. We formally tested that and could actually show that if maternal antibodies get transferred you need periodic exposure to this infection to actually build up your acquired or active immunisation. If you don’t do that you basically die of the infection. So I can summarise here. Memory is a nice idea. You can publish wonderful papers in Journal of Experimental Medicine. It’s earlier and higher, everybody can repeat that. But it doesn’t explain protection by vaccines or infectious agents. This is always antigen dependent. To keep up the protection you need re-exposure from within, from without, from antibody, antigen complexes. Now I’ve talked about the problem of T cell immunity and the immunopathology in such diseases as HIV, hepatitis B. There is also immunopathology going on for TB. And that’s why the granuloma is such a successful isolation procedure, to isolate the infection in a chronic infectious environment. And it is this granuloma configuration that keeps the antigen you know going on the middle. Without causing open or destructive TB, which of course when your T cell and protective immunity dwindles, be it by HIV and so on, then of course this control is not any longer functioning. Now let me conclude in general. Of course research to find out how things work is what the younger part of this audience is all here. And I think that’s as important and as pleasurable as sports and arts. Hard work, good environment and a good support and, particularly, good luck you know is the foundation. Open critical competition of course helps because the best critic is always from ourselves. Because if we can negate or falsify our own wishes, I think, we are safe from the nasty reviewers. Beliefs of course don’t help. Because I’ve had many good ideas, but ideas are cheap. You have to work to falsify them. And we should never make false promises. So 25 years ago HIV vaccine was promised within 2 or 3 years. We still don’t have it. And I hope I have shown you why. The virus is variable, has to persist, to keep up the antigen-driven protective activation of the T cell response. And of course we cannot imitate that yet - I don’t say it’s impossible. But this apparently has been successfully done by TB. It persists, keeps up the immune response but doesn’t kill you. Only after 60 but we said 22 is good enough. And HIV basically the same. Thanks very much for your attention. Applause.

Rolf Zinkernagel (2014)

Why Do We Not Have a Vaccine Against HIV or Tuberculosis?

Rolf Zinkernagel (2014)

Why Do We Not Have a Vaccine Against HIV or Tuberculosis?

Abstract

Analysis of the immune system is fascinating and progressing rapidly. As a field of medical enquiry, it has however, drifted and turned purely academic. This is because interest and appreciation of protective immunity in infectious disease medicine has been overtaken by ‘l’art pour l’art’ of so-called ‘basic immunology’. This development deprives much of immunological sciences of the biological basis and understanding that is linked to co-evolution of infectious agents and hosts’ protective immunity. It is this co-evolutionary context that renders this field so different from studying yeast, bacteria, fibroblasts, lymphocytes or neuronal cells in splendid isolation in in vitro model situations, where everything is possible (and permitted or mistakes forgiven without repercussions) because the co-evolutionary context is ignored.

I shall explain why we have excellent vaccines against acutely lethal (childhood) infections but not against most slow, chronic persistent infections or tumors, which usually kill us late i.e. after reproduction. Another conclusion is that so-called ‘immunological memory’ of course exists, but only in particular experimental laboratory models measuring ‘quicker and better’ responses, which often do correlate with, but are not the key, mechanisms of protection. Protection depends on pre-existing neutralizing antibodies or pre-activated T cells at the time of infection. This is well documented by the importance of maternal antibodies at birth for survival of offspring. Importantly, both high levels of antibodies in mothers are driven by antigen reencounter. This of course has serious implications for our thinking about old, and hopes for new, vaccines.

Further readings:

1. Zinkernagel RM, Ehl S, Aichele P, Oehen S, Kundig T and Hengartner H (1997) Antigen localisation regulates immune responses in a dose- and time-dependent fashion: a geographical view of immune reactivity. Immunol Rev 156:199-209
2. Zinkernagel RM (2001) Maternal antibodies, childhood infections, and autoimmune diseases. N Engl J Med 345:1331-1335
3. Zinkernagel RM and Hengartner H (2004) On immunity against infections and vaccines: credo 2004. Scand J Immunol 60:9-13

Content User Level

Beginner  Intermediate  Advanced 

Cite


Specify width: px

Share

COPYRIGHT

Content User Level

Beginner  Intermediate  Advanced 

Cite


Specify width: px

Share

COPYRIGHT


Related Content