Figure 1.
VEH and group A-like var expression levels among isolates from clinical and asymptomatic infections Presented are pie charts summarizing the var expression profiles of all isolates from children with clinical malaria (grouped into severe and non-severe malaria) and those with asymptomatic infection.
Each slice of a pie chart (whether blue or green) represents a unique var sequence with green representing group-A like var sequences. The size of the slice represents the percentage contribution of that sequence to the entire var expression profile of the isolate. The pie charts are ordered from left to right by increasing VEH and top to bottom by increasing group-A like var expression levels.
Figure 2.
Comparisons of VEH, IE surface antibody breadth, parasitaemia and host age.
The Mann-Whitney U test is used to compare asymptomatic and clinical infections in relation to VEH (A); heterologous IE surface antibody breadth measured as the median of the mean fluorescence intensities measured against eight different parasite isolates as described previously [21] (B); parasitaemia (parasites×105/ µl blood, C); host age (D). Blue arrows are used to indicate two individuals (46C and 46D) who became symptomatic with within the two days of sampling.
Figure 3.
The relationship between VEH and host IE surface antibodies Scatter plots are shown of the relationship between VEH and the breadth of the heterologous IE surface antibody response (A, see legend to figure 2); (B) shows the relationship between VEH and host age.
Spearman’s rank correlation coefficient rho was used to test for correlation.
Figure 4.
The relationship between VEH and host IE surface antibodies to heterologous and homologous isolates.
A-C show scatter plots of the relationship between VEH and breadth of the heterologous IE surface antibody response as in Figure 3A, but split into asymptomatic (A), non-severe (B) and severe malaria (C). (D) Shows the relationship between VEH and the homologous IE surface antibody response to each child’s own parasites among children with malaria (measured as mean fluorescence intensity). Among children shown in (D), (E-F) compares the heterologous and homologous responses among those with non-severe (E) and severe (F) malaria, respectively. The size of the marker in (E-F) is proportional to VEH. Spearman’s rho was used to test for correlation.
Figure 5.
Modeling the relationship between VEH and host IE surface antibodies.
In (A) we use pie charts to represent two modes in which acquired immunity may modify the parasite population in a single host. In “mode 1” IE surface antibodies alter the relative abundance of parasites expressing distinct PfEMP1 groups (pie chart slices, colours blue and green). In “mode 2” the acquired IE surface antibody repertoire is able to control a wide range of PfEMP1 variants such that only a small number of variants dominate the PfEMP1 expression profile of the infecting parasite population at any one time. Presented in (B) is an output from the mathematical model by Recker et al. This model was originally used to emphasize how a combination of long-lived antibodies to unique major epitopes and transient antibodies to shared minor epitopes can limit the number of synchronously expressed variants and prolong an infection [19]. Here, we replaced the duration of infection with a measure of the mean VEH of the parasite population following runs of the model with different initial levels of antibodies to major and minor epitopes. The aim was to simulate infections at different stages of their development. The graph shows the relationship between the initial number of each type of antibody [19] and mean VEH. We speculate that the ability of the host to induce transient partially cross-reactive antibodies in the model is equivalent to the observed breadth of IE surface antibody responses described in our dataset. Error bars represent the standard deviations.