RBCs are not homogeneous packets of parasite resource and differ globally (the haemoglobinopathies such as sickle cell anaemia, the thalassemiae and ovalocytosis), locally (e.g. blood groups) and within an individual (RBCs mature over time and when old, are cleared by the spleen). During a Plasmodium infection, the blood environment changes rapidly; in a naive host only a small fraction of the RBCs are reticulocytes (young RBCs), but which increases rapidly with malaria-associated anaemia. Several Plasmodium spp. have seemingly developed sophisticated mechanisms for adjusting to this variable environment and maintain the infection (Snounou et al. 2000). For example, in P. yoelii (a rodent parasite), the daughter merozoites of a single asexually replicating mother cell (schizont) expressed different variant proteins dictating RBC preference (Preiser et al. 1999). Such phenotypic plasticity in RBC preference would enable the parasite to rapidly adapt to the heterogeneous nature of the blood environment, whether due to its own action on the host or as a result of previously established parasites of the same or other species (Fig. 2). Several other multigene families are implicated in parasite survival. Plasmodium falciparum expresses variant proteins on the surface of the infected RBC (reviewed in Kyes et al. 2001) that enable parasite sequestration, thus avoiding destruction in the spleen. The expression of these adhesion proteins on the RBC surface exposes them to immune recognition. Within a single parasite genome there can be more than 50 copies of the var gene encoding for the variant proteins. However, any one parasite expresses only a single variant. Antigenic variation in clonal populations results in the switching of these variant proteins, such that in each asexual generation (48 h) only a small fraction of parasites exhibit a new variant and thus avoid immune targeting. This enables the parasite to persist but the switching is surprisingly low (order of few percentage express a novel variant each generation). Why risk sacrificing 98% of the parasite population to immune recognition and destruction each generation? One explanation is that as well as enabling sequestration and avoidance of the spleen, the immuno-dominant nature of this protein serves as a death flag, which results in auto-regulation of parasite densities and avoidance of runaway exploitation of the host (Saul 1999).
Figure 2. Parasite strategies responding to changes in vertebrate host blood environment as infection progresses. Strategy concerning red blood cell (RBC) preference: phenotypic plasticity in individual capacity (parasite with variant protein a or b or c) to invade RBCs of differing ages (mature vs. reticulocytes). Strategies governing sex allocation: total allocation in sexual gametocytes and allocation into male vs. female. At the beginning of an infection there is little anaemia, low production of erythropoietin hormone (Epo) and few reticulocytes. Parasites expressing variants for reticulocytes (b and c) do not find suitable RBCs. There is low investment in gametocytes and males particularly. As the infection progresses, there is increased anaemia, Epo and reticulocytes. Parasite variants favouring reticulocyte invasion proliferate, there is increased gametocyte production, especially in males to assure fertilization despite an increasing host antibody response. Epo is directly or indirectly involved in sex allocation.
Download figure to PowerPoint
Self-restraint, or prudent host exploitation, is contrary to current evolutionary thought on host–parasite interactions, largely because mutant unrestrained parasites would spread rapidly through the population. Such an eventuality may almost certainly be expected to arise for the human malaria parasites, P. falciparum and P. vivax, that cause 500 million new infections every year and generate several million parasites in each case. Indeed, as discussed above, there is some evidence that malaria parasites employ host exploitation strategies that maximize transmission success; by contrast, there has been no formal approach to the evolution of self-restraint in malaria parasites. However, there are several striking features of Plasmodium life history traits concerning the blood stage infection phase that are suggestive of self-restraint. Plasmodium spp. exhibit differential preferences for RBCs of differing ages. In the human parasite species, for example, P. vivax and P. ovale have a predilection for reticulocytes, P. malariae for mature RBCs and P. falciparum for all types. Similar variability in RBC preferences is found among all Plasmodium spp. Plasmodium spp. with reticulocyte predilections will initially grow slowly. By the time parasite-induced anaemia has induced increased reticulocyte densities, and thus potential for more rapid asexual replication, the host immune response would be expected to control the parasite. Those parasites targeting mature RBCs will be expected to rapidly deplete their RBC source and thus density will be checked by the RBC turnover. Parasite spp. with more catholic tastes, such as P. falciparum, abnegate resource-mediated control. Notably, however, the variant proteins enabling host exploitation via avoidance of the spleen (e.g. P. falciparum) or variant RBC specificity (e.g. P. yoelii) are highly immunogenic. Thus, although several parasite spp. have developed mechanisms to prolong infection and increase host exploitation, this facility is apparently accompanied by the induction of a strong immune response. This seems to go against the grain, as Plasmodium is known for the immune evasion structure of other surface molecules. Is it therefore possible that such antigenic variation simultaneously incorporates a self-restraint mechanism? Thus, as well as evading the immune response or adapting to the RBC population, does antigenic variation serve the pleiotropic role of reducing over-exploitation to maintain an infectious reservoir? Formal proof that such pleiotropic effects could evolve and be maintained is lacking. However, there is some empirical data that suggest this area warrants further thought. A selected lethal clone of the rodent parasite P. yoelii capable of invading all RBCs was restricted to reticulocytes in the presence of antibodies against this variant protein, and both the host and the parasite survived considerably longer. Similarly, under such a self-restraint perspective, reticulophilic parasites should generally be expected to invoke a greater immunological response to avoid runaway exploitation of the growing reticulocyte population (resulting from host response to parasite-induced anaemia). Comparative data is lacking, but P. vivax, for example, does elicit a strong fever response at a lower parasite density than P. falciparum (Boyd 1949). Finally, it should be remembered that the transmission success of the parasite does not solely rest on its interaction with its host vertebrate; the parasite needs the presence of the insect vector. It is thus feasible that optimum schedules of host exploitation will vary significantly between areas where the insect vectors are permanently active and where activity is highly seasonal. For example, prolonging the infection in the vertebrate host would be beneficial in regions of seasonal mosquito activity, in order to maintain an infectious reservoir. Despite considerable progress in our understanding of transmission from vertebrate to vector (as discussed next), the role vector population dynamics may play in shaping parasite host exploitation strategies remains unknown but potentially fruitful.