Research
Study Systems
Current
Previous
Sociality in malaria infections
Ecological interactions between parasites proliferating within a host can determine how sick a host gets, and how infectious. In-host ecology is also a potentially potent source of selection on pathogens, impacting on the evolution of virulence and drug resistance, and more generally, determining the fate of novel mutants. There are very clear parallels with social evolution here. Pathogens are evolving within a group structured environment, so that issues of sociality, relatedness and the strengths of within- and between-host selection become critical. Theoretical models of these processes abound; there is an alarming dearth of relevant experimental data.
Our work on in-host competition focuses on four questions:
- How does competition between malaria parasites work?
- What determines the outcome of competition?
- How does it affect clone fitness?
- How does it affect the evolution of virulence and other parasite traits?
We currently track the performance of individual clones in mixed infections using quantitative PCR of blood- and mosquito-stage parasites.
Group members involved: Andy Bell
Collaborators: Richard
Carter, Joanne Thompson, Hamza
Babiker.
Background and Open Questions
Many authors have pointed out that natural selection favours higher levels of virulence in genetically diverse infections because mild parasites will have lower fitness than more aggressive parasites. Even if host life expectancy is reduced so that all parasites do worse, mild parasites do disproportionately worse and are thus eliminated by natural selection. This can occur either through direct reductions of clonal densities in the presence of other clones (competition*) or, even in the absence of competitive suppression, if more virulent clones reduces the relative fitness mild clones by killing shared hosts. The direction of virulence evolution can be affected by the precise way in which clones interact in hosts. For instance, infection-blocking vaccines may lead to the evolution of reduced virulence if superinfection (competitive exclusion) occurs.
In the field, malaria infections frequently consist of more than one clonal lineage and a variety of indirect evidence is consistent with competitive interactions between them. However, it is almost impossible to demonstrate competition conclusively from non-experimental data. Using P. chabaudi, we have found, as have others, compelling evidence for strong within-host competition: clonal population sizes are reduced when other clones are present. But:-
(a) Clones can transmit as well or better from mixed-clone infections than from single infections (Taylor et al. 1997, Read and Taylor 2001). Thus, in-host competition can have no effect on parasite fitness or, paradoxically, enhance it. This unexpected finding is opposite to that assumed by virulence theory. However, these data concern two clones of similar virulence, and while we have seen the same phenomenon in other two-clone infections (Read et al. 2002), we failed to find it in a single experiment with three clones (de Roode, Read, Chan & Mackinnon, 2003). We need to see how general the phenomenon is and, importantly, how the in-host competitive ability of individual clones relates to their (i) virulence, and (ii) transmission (between host fitness). These relationships are central for understanding how in-host diversity affects virulence evolution, yet the few data we have on them in any disease system often contradict existing theory (Read and Taylor 2001).
(b) A range of initial conditions affects the outcome of competition within infections. In particular, infection order and, for coinfection, relative proportion in the inoculum, play a role. How do other factors, such as host genetics and semi-immunity (including strain-specific immunity against one of the competitors), affect the outcome of competition, both in the vertebrate and in terms of transmission success? We are also asking whether the determinants of clone fitness in two-clone infections are similar for infections consisting of three or more clones. If it is possible to formulate general rules relating transmission success, competitive ability and virulence of constituent clones, then it should be possible to predict the impact of a wide range of interventions on virulence evolution in situations where mixed infections are common.
(c) In our rodent model, aggregate (total) virulence is a function of the virulence of constituent clones. In mixed infections of clones with similar or with very different virulence, total virulence was maximal when genetic diversity in the inoculum was highest. This may reflect the costs of fighting a more diverse infection. However, at some inoculum frequencies, avirulent clones in an infection can reduce the virulence of virulent clones. Thus, depending on how mixed infections arise, host health can be enhanced or reduced. Intervention measures that reduce rates of superinfection could thus impact on virulence evolution in a manner analogous to vaccination. We are tracking to clones differing in virulence to determine how their frequencies in mixed infections act to generate aggregate virulence.
The details of in-host ecology are also relevant for understanding other evolutionary outcomes. The transmission consequence of competition is a key determinant of the spread of drug resistance, and the way in which antigenically distinct lineages compete will determine optimal rates of antigenic variation. It would also be informative to determine experimentally how the number of clones per host is related to disease severity and infectiousness. Diversity and severity are related in some but not all natural malaria populations, but both positively and negatively (review, Read and Taylor 2001). An otherwise ineffective vaccine reduced the number of clones per host in immunised people and, across populations, reductions in the force of infection, a major aim of many control programmes, is associated with reductions in clone number. The epidemiological and public health consequences of such reductions are currently very hard to predict.
**We use 'Competition' in the ecological sense to include resource-based interactions, direct attack, or reductions in clonal density by immune responses triggered by coinfecting clones (apparent competition). Including immunity as a subset of competition is somewhat unconventional in the disease context, but follows from the ecological definition (Read and Taylor 2001).