The evolution of virulence and ‘virulence management’

By definition, parasites/pathogens harm their hosts.  The degree to which parasites harm their hosts is called virulence.  The more virulent the parasite, the more it damages the host.  From an ecological perspective, we measure virulence in terms of reduced host survival and/or reproduction.

Why are some parasites/pathogens more virulent than others?  When should parasites evolve to be very virulent, and when should they evolve to be more benign?

Ewald (1993 and elsewhere) argued that a parasite’s virulence evolution should be related to the transmission mode of the parasite.  For instance, consider three types of transmission: direct transmission where the parasite cannot survive for long in the environment (e.g., the common cold – rhinovirus), direct transmission where the parasite can survive for a long time in the environment (e.g., smallpox), and vector transmission (e.g., malaria).  Virulence should be maladaptive if it hampers transmission – as in case 1.  If you’re so sick with a cold that you cannot leave your house, the virus is less likely to infect new hosts.  Virulence should be high whenever it increases transmission – as in 2 and 3.  If the virus can survive for a long time in the environment, why not go crazy replicating in the host (to the demise of the host) and then hang out in the environment until another host comes around?  And if the virus is vector-transmitted, why not replicate to a high density (to the detriment of the host) to insure that the vector gets a good dose of parasite with a blood meal?  Also, making the host lethargic might increase the likelihood that a vector gets a blood meal.

Ewald (1993) also suggested that by understanding how transmission can affect virulence evolution, medical scientists might be able to manipulate the evolution of virulence in important parasites/pathogens.  By reducing the probability of transmission, we could increase the cost of virulence (assuming that there is a trade-off between virulence and transmission).  Parasites should be more prudent when the probability of transmission is low.

Ewald (1993) gives several examples of pathogens that have evolved to become less virulent when the probability of transmission was reduced, but let’s just talk about HIV.  HIV is the sexually-transmitted retrovirus that causes AIDS.  HIV hangs out in white blood cells and can remain latent (=inactive) for long periods within the host, but the virus can also rapidly reproduce.  The longer the latent period, the less virulent the virus is, and the faster the reproduction rate, the more virulent the virus is.

How could we increase the probability of transmission of HIV?  If a human population were to change culturally from a monogamous, family-oriented culture to a more polygamous one, the rate of partner change and thus HIV transmission would increase.  Correspondingly, Ewald (1993) discussed some evidence that HIV can be more virulent in urban areas with many unmarried individuals than in rural areas that are more family-oriented.  (Of course, this is just one of many factors that affect the probability of HIV transmission, so please don’t go crazy on the cultural interpretations.)

How could we decrease the probability of transmission of HIV?  Condoms and safe sex education!  Ewald (1993) presented some evidence that the use of the drug AZT to treat HIV couldn’t completely explain the evolution of lower HIV virulence in homosexual males in urban areas in the 1980’s.  Increased use of safe sex practices might explain some of that decreased virulence.

Ewald (1993) is a bit old – there’s some great evidence for/against this idea of ‘virulence management’ in the literature now.  But the Ewald (1993) paper is a good read and a ‘popular’ article, so check it out!


Ewald, P.W. 1993. Evolution of virulence.  Scientific American.

Related 2001 interview with Dr. Ewald from PBS.

T-cell Vaccines and Host Pathology

This is one of several posts that I’ll write about the Ecology and Evolution of Infectious Disease conference (EEID).  I’m starting with the very first talk at the conference, because I’m not feeling particularly creative, and chronological order is my default.  🙂

Let’s talk about vaccines.  The vaccines that you’re familiar with are probably vaccines that cause antibody immunity.  For instance, you might be given a bit of dead pathogen, and your immune system learns how to make antibodies that target the pathogen.  For instance, vaccines for influenza, chickenpox, polio, and hepatitis B are meant to help you develop antibody immunity.

But what if the pathogen mostly hangs out in your cells, and antibodies won’t cut it?  This happens with many pathogens that have persistent infections, like HIV and malaria.  As you know, we don’t have vaccines for malaria and HIV, because these antibody-type vaccines won’t work (yet? ever?).

People are working on a different type of vaccine, called a T-cell vaccine.  T-cells are immune cells that find and kill host cells that are infected with pathogens.  The hope is that T-cell vaccines will better enable T-cells to recognize infected cells.

But here’s the problem.  Some people have found that vaccines for these pathogens that hang out inside of cells can actually cause disease/pathology.  That is, individuals (e.g., mice) have greater pathology when they are exposed to a pathogen after being vaccinated than if they hadn’t been vaccinated.  (Note:  I’m talking about vaccines that are still being developed, not ones that we currently use.)  Obviously, we can’t develop useful vaccines until we figure out why they may be increasing infection and causing pathology.

To work on this, Johnson et al. (2011) made a model (which parallels experimental data) of lymphocytic choriomeningitis virus in mice.  What they found was that at low and high densities of T-cells, pathology was low.  But at intermediate levels of T-cells, pathology was high.  In this case, “pathology” was measured in terms of the number of functional host cells, where few functional host cells means high pathology.


My cartoon of Rustom Antia’s graph.

Why should the number of T-cells matter?  Rustom Antia suggested that at low T-cell numbers/density, the virus kills some host cells and the T-cells don’t kill many host cells.  At high T-cell numbers/density, the virus kills few host cells, but the T-cells kill more host cells.  In both cases, not too many host cells are killed.  At intermediate T-cell numbers/density, both the virus and the T-cells kill many host cells, which ends up being detrimental to the host.  Tada!

So, there you have it.  Now, since this isn’t my area of expertise, I recommend checking out the paper that Rustom based his talk on to find out more.

What do you think?  Will we figure out how to use T-cell vaccines in the near future?


Rustom, Antia. 2013. An immuno-epidemiological approach to understanding vaccine efficacy. EEID.

Johnson et al. 2011. Vaccination Alters the Balance between Protective Immunity, Exhaustion, Escape, and Death in Chronic Infections. Journal of Virology 85(11): 5565–5570.