Cross-Species Interactions Between Malaria Parasites in Humans

See allHide authors and affiliations

Science  04 Feb 2000:
Vol. 287, Issue 5454, pp. 845-848
DOI: 10.1126/science.287.5454.845


The dynamics of multiple Plasmodium infections in asymptomatic children living under intense malaria transmission pressure provide evidence for a density-dependent regulation that transcends species as well as genotype. This regulation, in combination with species- and genotype-specific immune responses, results in nonindependent, sequential episodes of infection with each species.

In malaria-endemic regions, humans commonly harbor chronicPlasmodium infections consisting of complex mixtures of different species (1) and genotypes of parasites (2). Longitudinal studies of animal malaria infections have shown that infection dynamics are affected by cross-species immunity, resulting in within-host interactions between species [reviewed in (3)]. Direct evidence for the action of cross-species immunity in human malaria infections has been lacking. Consecutive experimental infections with different species and genotypes indicated that immunity to human malaria is species- and genotype-specific (4). Data on the dynamics of simultaneous multispecies (5) and multigenotype coinfections (6) are available from only a few experiments. In some instances of mixed infections of P. falciparum andP. vivax, replacement of one species with another has suggested that the species can interact. However, the relevance of such data to natural infections in individuals living in endemic regions is unclear. In such regions, individuals are superinfected from birth, whereas experimental data are derived from primary infections in nonimmune adults.

Indirect evidence for interactions between humanPlasmodium species comes from cross-sectional malaria surveys in which there is a deficit of mixed infections relative to that expected assuming no interaction (7), reciprocal seasonality in the prevalence of different species (3, 8), and reduction in the severity of malaria symptoms in individuals with limited pre-exposure to different species (9). None of these studies has provided information about parasite dynamics. To investigate the possible role of species-transcending immunity in the control of malaria parasitemia in humans, we analyzed the dynamics of multiple, coinfectingPlasmodium species and genotypes in infected but clinically asymptomatic children.

The intrahost dynamics of Plasmodium species were determined in 34 children aged 4 to 14 years, resident in Papua New Guinea (10). The children were exposed to all four species causing human malaria (11) at an estimated rate of 0.86 infectious bites per person per day (12). Parasites were sampled every 3 days for 60 days and species density was quantified by microscopy (13). Genotypes of P.falciparum and P. vivax were characterized by molecular typing (Table 1) (14).

Table 1

Microscopy, genotyping, and statistical results. The Bulmer test (18) was not used on data from 13 children with <50% positive smears and two children with autocorrelated total parasite density data (24). “nd” indicates not done. Dashes indicate where a species was not detected. Pf, P. falciparum;Pv, P. vivax; Pm, P. malariae; Po, P. ovale.

View this table:

Parasites were present in all 34 children (Table 1) and 82% were infected with more than one species. Many were infected with multipleP. falciparum and P. vivax genotypes (Table 1) (14). The number of genotypes was positively correlated with the proportion of positive smears across all age groups for P. vivax (Fig. 1A) (r = 0.709, P = 0.002) and in children aged 5 to 14 for P. falciparum (Fig. 1B) (r= 0.742, P = 0.0001). These correlations demonstrate that increased smear positivity for each species was due to more infections and not maintenance of a single genotype. Persistence of single genotypes was observed only for P. falciparum in children aged 4 years (Fig. 1B) (15).

Figure 1

Plot of number of genotypes of (A) P. vivax and (B) P. falciparum per child against the proportion of smears positive under microscopic analysis for each species.

Children with the highest proportion of smear-positive samples (Table 1) exhibited relative stability in total parasite density despite complex underlying dynamics attributable to each species (Fig. 2A). The dynamics of individual genotypes were also highly variable (15). The density of each species changed over time, while total parasitemia was regulated around 1000 parasites/μl of blood, for periods longer than any single episode (16). This value is close to fever thresholds for children in this population (17) and is in agreement with the absence of clinical symptoms (10).

Figure 2

Parasite dynamics in (A) child 19 and (B) child 31. Numbers between panels indicate sampling day over the 60-day period. (Upper panels) blue bars show total parasite density; light gray shading indicates the lower microscopy sensitivity level (40 parasites/μl); dark gray shading indicates the fever threshold [1000 parasites/μl (17)]; open bars are smear-negative samples. (Lower panels) Proportion of each species is shown for P. falciparum (orange), P. vivax (green), and P. malariae (yellow).

We tested total parasite density data for density-dependence using Bulmer's test (18), which distinguishes random changes from a tendency to return to an equilibrium value. Evidence for density-dependence was detected in log10 total density data from all but 3 of 19 children tested (Table 1). Significance was reduced when the test was applied to each species, in 9 of 13 children with sufficient data to test, indicating the cross-species nature of regulation. Where a single species predominated but multiple genotypes were present (children number 6, 31, 4, 32, and 14), total parasitemia also exhibited density-dependence (Fig. 2B and Table 1), suggesting that single-species coinfections are regulated similarly. Maintenance of total parasite density around a threshold is inconsistent with dynamics resulting only from species- and/or genotype-specific immunity responsible for parasite clearance [reviewed in (19)]. To account for these dynamics, additional species-transcending, density-dependent regulation of parasite density is required.

To determine if there was evidence for interactions between species, we analyzed the patterns of infection in the children. The number of children infected with >1 species during the study was no different from that expected, assuming independence of each species (χ2 = 1.60, 2 df, P = 0.449). In contrast, a deficit of mixtures of species was detected when data were analyzed cross-sectionally (χ2 = 9.06, 2 df,P = 0.028). The peak density of episodes of different species (16) did not coincide as often as expected by chance (χ2 = 6.56, 1 df, P = 0.010). Together, these results demonstrate that episodes tended to follow a sequential rather than a concurrent pattern.

Sequential infections with different Plasmodium species have been reported from a few other longitudinal data sets [reviewed in (1)], but these were not interpreted as the effect of density-dependent regulation. In symptomatic, nonimmune individuals, fever may underlie the stability of parasitemia (20), but in these single infections the Plasmodium immune evasion mechanism of antigenic variation (19) could also explain these dynamics (21). Fever prevalence in children aged 4 to 14 in Papua New Guinea is insufficient to fully account for our data (17). Neither antigenic variation nor density-dependent regulation alone can explain both the stability of parasitemia and sequential infections observed in our data.

We believe that the species interactions result from the interplay between density-dependent regulation and the differential growth and clearance rates of individual parasite populations resulting from clonal antigenic variation (19, 22). Growth of one parasite population to above threshold density will trigger density-dependent regulation. When this happens, minority coinfections will also be inhibited. When the majority population is cleared by species- and/or genotype-specific responses (19,22), total density will drop below the threshold and density-dependent regulation will cease. In these circumstances, minority populations could expand because density-dependent constraints would be absent. In this way, the sequential episodes of infection we have observed can be generated.

If species-specific vaccines [e.g., (23)] are successful in reducing or preventing P. falciparum parasitemia, the effect in highly endemic regions could be to provide an increased opportunity for other species to multiply when the constraint imposed by density-dependent regulation is removed. This could result in increased prevalence, transmission potential, and disease unless vaccines include targets specific for other species.

  • * To whom correspondence should be addressed. E-mail: marian.bruce{at}


View Abstract

Stay Connected to Science

Navigate This Article