S. Gandon2
S. Nee1
J.C. de Roode1
M.J. Mackinnon1.
1Institutes of Evolution, Immunology and Infection Research, School of Biological Science, University of Edinburgh, Edinburgh, UK.
2CEPM, UMR CNRS-IRD 9926, IRD, Montpellier, France
Marek’s disease has become more virulent since the Second World War, and there is very good evidence from the work of Dick Witter and others that is because of genetic change in the pathogen population. We use the term ‘virulence’ to mean ‘disease severity’.
Here we discuss possible mechanisms by which natural selection would favour more virulent pathogens, and in particular, the possibility that vaccination could prompt evolutionary increases in virulence. Vaccination could drive MDV virulence upwards, but testing this hypothesis requires detailed experiments which have yet to be done. We illustrate the sort of thing that is required by reference to our experimental work on rodent malaria.
There is a large body of theoretical work in evolutionary biology examining the evolution of virulence. We follow the classical pathogen trade-off model. The idea is that pathogen strains within a species range on a virulence spectrum from benign (only weakly symptomatic) through to virulent (highly lethal). Selection is then assumed to act on this variation as follows. Excessively virulent pathogen mutants are eliminated by natural selection because they kill their hosts and therefore themselves. Excessively avirulent strains also have low fitness because they are more rapidly cleared from their hosts or they fail to maximise their output of transmission profiles. Natural selection should therefore optimise the balance between the costs of virulence (host death and/or morbidity) and the benefits (immune evasion, host resource extraction and hence transmission). Empirical support for this trade-off framework comes from a number of animal and plant models (for a lead into this literature, see Mackinnon and Read 2004a and Read et al. 2004).
Could vaccination alter the fitness costs and benefits of virulence, and hence the level of virulence favoured by natural selection? We have recently argued that it could (Gandon et al. 2001, 2003; see Read et al. 2004 for a possibly more digestible version of these arguments, and Mackinnon and Read 2004a for review of these ideas as applied to malaria). The logic is that vaccination is used to prolong host life expectancy. But while vaccines protect hosts, they also protect more virulent strains from killing their hosts and hence themselves. Thus, vaccination relaxes selection against virulence, so that more virulent strains circulate in a population. Non-immune individuals are thus at great risk of death.
This argument is a very general one, but in a model for year round endemic malaria, we showed (Gandon et al. 2001) that this evolution always erodes away the public and animal health benefits of vaccination, making the overall disease burden higher than in that seen in a clinical trial. There are some regions of parameter space where this evolution would make things worse than in the pre-vaccine era (ie. more people would die of malaria). This evolution could be very quick (Gandon et al. 2003) or a matter of a few decades (Gandon et al. 2001). Not all vaccines will prompt this evolution: vaccines which stop hosts becoming infected in the first place, or which stop hosts transmitting, do not have this effect because they do not affect the costs and benefits of virulence.
Vaccines which reduce symptoms either directly (anti-symptom or anti-toxin vaccines) or which reduce symptoms by reducing pathogen titres (anti-replication vaccines) without completely halting transmission (leaky vaccines) will favour the evolution of more virulent strains because they reduce the costs of virulence (risk of host death) while allowing virulent strains to pass through vaccinated hosts.
This argument assumes that (a) there is genetic variation in pathogen virulence on which selection can act, (b) vaccines are leaky, (c) the selection against virulence is from host death (or at some other pressure which vaccine relaxes), and (d) there are fitness advantages to virulence in the absence of host death. This last assumption is key: there needs to be upward selection on virulence for any level of virulence to persist. We have experimentally demonstrated fitness advantages to virulence in the model of rodent malaria we work with: more virulent clones transmit better (Mackinnon and Read 1999a), and have a competitive advantage in mixed clone infections (Mackinnon and Read 1999b, de Roode et al. 2004, in review). This is also true in immunised animals (Mackinnon et al. 2003, Mackinnon and Read 2004b). The other three assumptions likely also hold in malaria (Mackinnon and Read 2004a).
The general lessons we draw from this are as follows:
- Leaky vaccines are evolutionary disasters waiting to happen. Vaccines and other control measures should be aimed at stopping transmission as well as reducing disease severity.
- Genetic resistance could have the same effect. So selecting for enhanced resistance in agricultural will also prompt the evolution of more virulent strains, where the resistance works in a manner analogous to a leak anti-pathogen or anti-symptom vaccine.
Have these processes caused the virulence increases in MDV? Our reading of the literature is that it is currently impossible to be sure. However, we note that a number of lines of evidence point this way. More virulent MDV strains appear first in vaccinated flocks. MDV vaccines leak. MDV vaccines are anti-pathogen and possibly anti-toxin (anti-symptom) vaccines. More virulent MDV strains do better in vaccinated hosts. More virulent strains are not escape mutants; they are antigenically identical to less virulent strains.
To our minds, the key unresolved issue in MDV is how virulence relates to transmission (fitness).
Without knowing this, it is impossible to address the following key questions: (1) are more virulent strains more infectious if the host does not die? (2) What selected against more virulent strains in the pre-vaccine era? (3) Does vaccination reduce the fitness costs of virulence? (4) Do vaccines drive the evolution of virulence?
These questions could be answered with experiments analogous to ours on malaria, summarised above. In particular, though, we think there is substantial potential for serial passage experiments through immunised and non-immunised chickens.
Does immunity promote the evolution of more virulent strains, as we have shown for malaria (Mackinnon and Read 2004b)?
We think the resolution of this question is important to the poultry industry, and more generally. If vaccination is responsible for increasing MDV virulence, why could the same not occur in human diseases? Alternative control options are considerably more limited in human populations than they are in the poultry industry.
References
Ferguson, H. M., Mackinnon, M.J,. Chan, B.H.K., and Read,A.F. 2003. Mosquito mortality and the evolution of malaria virulence. Evolution 57 (12): 2792-2804
Gandon, S., Mackinnon, M. J., Nee, S. & Read, A.F. 2001. Imperfect vaccines and the evolution of pathogen virulence. Nature 414: 751-756
Gandon, S., Mackinnon, M., Nee, S., & Read, A.F. 2003. Imperfect vaccination: some epidemiological and evolutionary consequences. Proceedings of the Royal Society of London Series B 270:1129-1136
Mackinnon, M. J., & Read, A.F. 1999a. Genetic relationships between parasite virulence and transmission in the rodent malaria Plasmodium chabaudi. Evolution 53: 689-703.
Mackinnon, M. J., & Read, A.F. 1999b. Selection for high and low virulence in the malaria parasite Plasmodium chabaudi. Proceedings of the Royal Society of London Series B 266: 741-748.
Mackinnon, M.J., & Read, A.F. 2003. The effects of host immunity on virulence-transmissibility relationships in the rodent malaria Plasmodium chabaudi. Parasitology 126: 103-112.
Mackinnon, M.J., & Read, A.F. 2004a. Virulence in malaria: an evolutionary viewpoint. Philosophical Transactions of the Royal Society of London. Biological Sciences 359, 965-986.
Mackinnon, M.J. & Read, A.F. 2004b. Immunity promotes virulence evolution in a malaria model. PLoS Biology 2 (9) e230 DOI: 10.1371/journal.pbio.0020230.
Read, A.F., Gandon, S., Nee, S., & Mackinnon, M.J. 2004. The evolution of pathogen virulence in response to animal and public health interventions. In: Dronamraj, K. (ed.) Evolutionary Aspects of Infectious Diseases. pp. 265-292. Cambridge University Press.
de Roode, J.C., Culleton, R., Cheesman, S.J., Carter, R., & Read, A.F. 2004. Host heterogeneity is a determinant of competitive exclusion or coexistence in genetically diverse malaria infections. Proceedings of the Royal Society of London Series B 271: 1073-1080.
de Roode, J.C., Pansini, R., Cheesman, S.J., Walliker, D & Read, A.F. (in review) Virulent parasites are competitively superior in genetically diverse malaria infections.
From Proceedings of the “7th International Symposium on Marek’s disease”, St. Catherine's College, Oxford, United Kingdom.
