Plasmodium falciparum genotypes, low complexity of infection, and resistance to subsequent malaria in participants in the Asembo Bay Cohort Project

Abstract

To assess the relationship between the within-host diversity of malaria infections and the susceptibility of the host to subsequent infection, we genotyped 60 children's successive infections from birth through 3 years of life. MSP-1 Block2 genotypes were used to estimate the complexity of infection (COI). Malaria transmission and age were positively associated with the number of K1 and Mad20 alleles detected (COI(KM)) (P < 0.003). Controlling for previous parasitemia, transmission, drug treatment, parasite density, sickle cell, and age, COI(KM) was negatively correlated with resistance to parasitemia of > 500/microl (P < 0.0001). Parasitemias with the RO-genotype were more resistant than those without this genotype (P < 0.0000). The resistance in low COI(KM) infections was not genotype specific. We discuss the impact of genotype-transcending immunity to conserved antigenic determinants. We also propose a diversity-driven immunomodulation hypothesis that may explain the delayed development of natural immunity in the first few years of life and suggest that interventions that decrease the COI(KM) could facilitate the development of protective immunity.

FIG. 1

Frequency of detecting MSP-1 Block2 alleles in 60 children's parasitemias from birth to 3 years of life. Using MSP-1 K1-, Mad20-, and RO-specific primers, 668 parasitemias were genotyped. There were 23 K1, 20 Mad20, and 1 RO. The frequency of detection was the number of samples (parasitemic) with the allele divided by the total samples tested. Parasitemias following a month of aparasitemia or an antimalaria treatment (isolated parasitemias) gave similar distributions (data not shown).

FIG. 2

Spatial and temporal EIR in Asembo Bay Cohort Project. Villages could be divided into three categories based on long-term EIR: low (N = 4), medium (N = 7), and high (N = 4). The means and upper SE are shown. Time periods when the average EIR were >0.5 infected bites/person/night (indicated by horizontal line) were defined as hypertransmission seasons. Season boundaries are indicated by vertical lines. J-93, January 1993, etc.

FIG. 3

Parasitemias of <500/μl were not associated with frequency of febrile illness (A) or moderate anemia (B). The routine monthly blood samples from 28, 185 ABCP children 0.0 to 0.5 (green), 0.5 to 1.0 (purple), 1.0 to 1.5 (red), 1.5 to 2.0 (orange), or 2.0 to 3.0 (blue) months old, born into the cohort, were used. Parasitemia density categories were made by binning the data. The frequencies were determined for each parasitemia group and age group. The youngest age group is only plotted in panel A due to the difficulty in assessing anemia in this age group. Parasitemias of <500/μl were not associated with increased probability of febrile illness or moderate anemia (P > 0.05 for each age group).

FIG. 4

COI_KM and RO parasitemia with the mean time of resistance to subsequent parasitemia of >500/μl and febrile parasitemia. We show the association, controlling for malaria transmission (A), administration of antimalaria drug treatment (B), absence of the RO parasite genotype (C), and sickle cell genotype (D). The results for individuals in the high-transmission and nontreated categories are stratified by age and plotted with SE bars. The parasitemias of <500/μl occurring within an interval of resistance were not genotyped and considered with regard to resistance to exclude overlap. For each stratification, n > 15. COI_KIM20 and RO (P < 0.0001) and sickle cell (P = 0.0031) independently associated with months resistant.

FIG. 5

Representative successive parasitemias and MSP-1 Block2 allele detection for five children (A to E). The allele lengths (K1 in red, Mad20 in blue, and RO in green) and parasite densities (black asterisk, not febrile; orange asterisk, febrile/treated) are shown. The dotted line = 500/μl. K1 alleles were often not detected in the subsequent 2 months. Children were examined in their homes every 2 weeks. Blood samples were taken at least every month. A reported or measured fever or illness resulted in an additional blood sample's being taken. Blood samples were collected during treatments (Fansidar administered at day 1, 3, and 7 postclinical malaria detection) and suggested effective parasite clearance. Periods of aparasitemia and treatment are not shown for clarity. The same K1 allele was rarely detected more than once in 2 months, but analysis showed that this was not significantly different from chance (P > 0.05 for any allele).