Experimental anti-malaria vaccine accidentally makes malaria worse :-/

The Evolutionary Consequences of Blood-Stage Vaccination on the Rodent Malaria Plasmodium chabaudi

The concept of vaccination is, superficially, simple-- Safely mimic 'infection' so your immune system learns how to fight a pathogen, without needing to get sick from the genuine pathogen.  Then if you are ever exposed to the real, scary pathogen, your immune system already knows how to deal with it.

But things dont always go according to plan...

In 1966, a novel vaccine against Respiratory Syncytial Virus was a tragic failure.  The lucky kids didnt respond at all.  The unlucky kids made a suboptimal antibody response, that ended up making RSV infection worse, ultimately killing two toddlers.

Something that has plagued HIV World in our attempts to generate an anti-HIV vaccine is called 'antibody dependent enhancement'.  Lets say you make a GREAT anti-HIV vaccine that generates lots of anti-HIV antibodies.  Hurray!!!!... right?  Well, those antibodies stick to HIV.  Macrophages stick to antibodies (Fc receptors).  HIV LUVS infecting macrophages... so... the 'anti-HIV antibodies' would actually be bringing HIV to its target, rather than protecting against infection.  :-/  If we dont make an HIV vaccine right, it could make things worse.

But these two examples show how the immune system and its response to vaccines could make things worse.

What this publication shows is how the pathogens response to the immune systems response to the vaccine could make things worse.

Did ya catch that? ;-D

Here is what happened-- Barclay et al made an anti-malaria vaccine.  They wanted to test it against 'the best' malaria (if your vaccine can stop 'the worst', it can stop the rest) so they artificially evolved an extremely pathogenic malarial variant, and gave it to mice.  Some of those mice also got the experimental vaccine, and HURRAY!  The vaccine, while not sterilizing, could help the mice keep the parasites in check-- lower parasite loads, higher red blood cell counts.  YAY!

But that 'not sterilizing' thing had the researchers wondering-- What does a malaria parasite that can escape their vaccine look like?  Great question!  Someone who got the vaccine could still be infected, which means another mosquito could cruise by and pick up vaccine resistant malaria, and infect someone else.  We need to understand how malaria was evolving in response to the vaccine (a selective pressure).  So to figure that out, they took some of the parasites from the vaccinated mice, and used them to infect another vaccinated mouse, over and over, until they got a few lines of malaria that could happily escape their vaccine.  Fantastic!  Evolve something in the lab before it evolves in Nature, so you know what to expect, you might be able to figure out how to fix the problem!

... Minor problem.  When they looked at the malaria that could escape their vaccine, they expected to see mutations in the genome that coded for the target of their vaccine.  Of course.  The target had to have changed to escape the antibodies induced by the vaccine, right?


There was no genetic difference in the target of the vaccine between the 'regular' and 'vaccine escape' malaria.

The mutations were elsewhere.

And those mutations created a MORE pathogenic version of malaria.  The 'vaccine escape' malaria didnt escape the vaccine, directly.  It looks like malaria just evolved to replicate more efficiently-- There were more parasites in the vaccinated mice, which lead to lower red blood cell counts (worse disease).

And here is where things get really tricky-- Usually, when things evolve to escape a selective pressure, it might appear to be 'more fit' in the presence of the selective pressure, but when you take that pressure away and compete the 'more fit' variant against a natural variant, it is actually LESS fit (example: drug resistant forms of HIV).  But in this case, the 'more pathogenic' malaria was actually more pathogenic, whether the mouse had gotten the vaccine or not.  And, in the examples I gave above (RSV, HIV), only the people who got the vaccine would be at risk of side-effects.  In this case, even mice that did not get the vaccine were more susceptible to the vaccine-evolved malaria.


On the bright side, there are a lot of caveats to this research (example: mice are not a very good animal model for malaria).  Its not like 'more pathogenic malaria' WILL be the result of ANY human anti-malaria vaccine.  And its not like this evolutionary pathway was easy.  The serial passages crapped out, sometimes. The vaccine did try very hard to do what it was supposed to do, so sometimes they had to go back to a previous passage and try again.

But of all the times they tried, it only took about ten passages for malaria to 'figure out' an answer to the vaccine.  No Edge of Evolution here.

So we dont have to be 'scared' because of the results of this paper-- we need to be aware.  We need to understand vaccines.  We need to understand evolution.

More like this

That's a great little write up and a very interesting read. Thanks!

By Adam Morrison (not verified) on 01 Aug 2012 #permalink

Once again, reality trumps fiction, and we discover that evolution is never a simple process.

A most intriguing write-up, ERV.

By Bob Powers (not verified) on 02 Aug 2012 #permalink

ERV = endogenous rhabdovirus?

By mo (one of Abb… (not verified) on 03 Aug 2012 #permalink

This is interesting, but it is not clear to me that this is a serious problem, or that it has much to do with vaccination per se. The “selection” that may be driving this behavior is the use of blood from mice at precisely 7 days post-infection to infect the next cohort.

Of course using blood at the peak of parasite density will select for strains with a higher initial growth rate, as the sham vaccinated passages also showed compared to the ancestral strain.

Since these are all clones, it is not clear to me that the change in phenotype is due to “evolution” per se (changes in genotype due to selection), it could be due to epigenetic programming changes; that is that the “virulence” of the transferred parasite depends on how long it was in the host prior to transmission, short time, highly replication rate, long time slower replication rate.

If you compare figure 2D and figure 3D, the difference between the sham and vaccinated passages goes down as the passage number increases from 10 to 21.

In figure 4C, the error bars of symbols of parasite density in unvaccinated mice overlap. The two curves are indistinguishable. The parasite density (selected and not selected) in vaccinated mice is always significantly lower than the parasite density in unvaccinated mice, with essentially no overlap.

If you look at figure S3, parasite density in vaccinated animals is always lower than in non-vaccinated animals. But this is peak parasite density, not average parasite density.

Mosquitoes don't only bite at the peak of parasite density. What is important for transmission is how many parasites get picked up and is an infective dose transmitted to the next organism. If you look at figure 1C, the density of parasites is always lower in the vaccinated animals.

With natural transmission due to mosquitoes, if the parasite density is lower, there should be less transmission. If the transmission is reduced to less than one new case per infection, then the disease dies out. That is the whole point of vaccination and herd immunity. The immunity doesn't need to be sterilizing (and it was not in this case), to prevent epidemics, it only has to reduce transmission enough so that the disease doesn't spread.

In other words, if now there are 1.5 new transmissions per infected individual and widespread vaccination drops that to 0.75 new transmissions per infected individuals then malaria would eventually die out.

By daedalus2u (not verified) on 03 Aug 2012 #permalink

Patrick, the antibodies that vaccines generate are coded for in the DNA of the organisms' immune system. Natural exposure to disease and exposure to antigens in vaccines both can only generate antibodies from immune system DNA.

Vaccination reduces the chances of tolerance over no vaccine because the number of exposures of the agent to the immune system is greatly reduced. In a full blown course of a disease, the levels of infective agent become very high as it is replicating internally. Each instance of infective agent replication is an opportunity for a mutation and an escape from the form that is susceptible to the antibodies the immune system is capable of generating.

If you reduce the number of possible mutation events by 99.9% by decreasing the pathogen load following infection, reducing transmission of the disease by making people immune and also by reducing transmission through herd immunity, the possible mutation and escape into something resistant to the immune system is much less. It is not like antibiotics.

If a virus did "escape" from the immune system (the way that flu does), all that means is that new vaccines are needed. That is accepted in flu because that is how flu works. If it happened in other diseases then they would require repeated vaccination every year the way that flu does.

By daedalus2u (not verified) on 06 Aug 2012 #permalink