Fine specificity of anti-MSP1₁₉ antibodies and multiplicity of Plasmodium falciparum Merozoite Surface Protein 1 types in individuals in Nigeria with sub-microscopic infection

The cross-sectional study was conducted in Igbo-Ora, a rural area in south-western Nigeria; approximately 100 km from the state capital Ibadan. In Igbo-Ora malaria is hyperendemic with a prevalence of approximately 75% in children during the raining season from April to October and 55% to 60% during the dry season from November to March. The peak incidence was observed in November and August, with lower rates observed in July, September and October [30].

Blood was collected by venipuncture into heparinized and EDTA containing tubes. Thick and thin blood smears were stained with Giemsa for microscopic detection of malaria parasites, examining 200 fields. Malaria parasites were counted relative to leukocytes, and a leukocyte count of 8,000/μl of blood was assumed. Blood slides were examined using the quality control procedures described elsewhere [30]. Red blood cells were separated from the plasma by centrifugation and both were stored at -80°C prior to further analysis.

Haemoglobin type was determined by electrophoresis on cellulose acetate gel with alkaline buffer. Packed cell volume (PCV) was determined using a haematocrit centrifuge (Hawksley and Sons, Ltd., Lancing, United Kingdom). Anaemia was defined according to the age-dependent PCV levels published by the WHO (0-3 months: < 32%, 6 months-4 years: < 33%, 5-11 years: <34%, >12 years: < 37%; for adults, male: < 40% and female: <37%).

DNA was extracted from blood cells using the QIAamp DNA blood mini kit (Qiagen) according to the manufacturer's procedure. Single and nested PCRs were performed to amplify the polymorphic sequence block 2 of P. falciparum msp1 as described previously [31]. Amplifications were performed in a final volume of 50 μl and cloned or monomorphic parasite lines were used as positive template controls. The first round PCR amplification was performed using two microlitres of DNA and the primers 5'-AAGCTTTAGAAGATGCAGTATTGAC-3' and 5'-ATTCATTAATTTCTTCATATCCATC-3', designed on the conserved regions in sequence blocks 1 and 3 on either side of block 2. The second round nested PCR amplification was carried out using 2 μl of the primary amplified products. Primers were used for detecting the K1 (5'-AAGAAATTACTACAAAAGGTG-3' and 5'-TGCATCAGCTGGAGGGCTTGCACCAGA-3'), RO33 (5'-AGGATTTGCAGCACCTGGAGATCT-3' and 5'-GAGCAAATACTCAAGTTGTTGCA-3') and MAD20 (5'-TGAATTATCTGAAGGATTTGTACGTCT-3' and 5'-GAACAAGTCGAACAGCTGTTA-3') msp1 block 2 sequence families [32]. The amplifications were performed in a Biometra Uno II thermal cycler (Biometra, Göttingen, Germany) and the products were loaded onto a 1.5% agarose gel (Peqlab Erlagen, Germany), electrophoresed and stained with ethidium bromide. The DNA was visualised under ultraviolet light and the fragment size polymorphism for each allelic family was determined. It was assumed that each PCR fragment represented at least one parasite genotype.

The recombinant Pf MSP1₁₉ protein (rMSP1₁₉) was prepared by standard techniques as a glutathione S-transferase (GST) fusion protein by expression from a pGEX-3X plasmid as described previously [33] and it represents the sequence of the Wellcome parasite line [34] (EMBL accession number X02919).

An indirect ELISA was performed to determine the prevalence and titre of IgG antibodies binding to rMSP1₁₉, as previously described [16]. Ninety six well-polystyrene microtitre plates (Costar, Corning Inc. NY) were coated with 0.5 μg/ml rMSP1₁₉. Following incubation with diluted plasma samples, bound antibodies were detected with horseradish peroxidase (HRP)-rabbit anti-human IgG conjugate (Dako, Denmark) and ABTS/hydrogen peroxidase substrate solution by measuring the absorbance at 650 nm. The plasma samples were diluted at a 1:25 ratio and then in two fold dilutions to 1:3200; the reciprocal end point titre (the highest dilution that gave an absorbance value greater than the background level measured with negative control samples) was log transformed, and data were expressed as geometric mean log reciprocal titres. The negative control samples were obtained from malaria-naïve residents of the United Kingdom.

The ability of polyclonal antibodies in the human plasma samples to compete with mAbs 12.8 and 12.10 for binding to rMSP1₁₉ was examined by competitive ELISA [22]. Initially each mAb was titrated against the rMSP1₁₉ and then used in the competition assay at a concentration that gave an absorbance reading just below the maximal absorbance for that antibody (i.e. just below the top of the titration curve). Microplate wells were coated with 0.1 μg of rMSP1₁₉, then plasma at a dilution of 1:50 or 1:250 was added to duplicate wells, and the plates were incubated overnight at 4°C. A predetermined amount of mAb was then added and incubation continued overnight; the amount of bound mAb was determined as described previously [22]. Plasma samples were divided into two groups: non-competitive samples, which were those that inhibited the mAb binding by less than 50% at a 1:250 dilution and competitive samples, which were those that inhibited the mAb binding by more than 50% at a 1:250 dilution.

During the enrolment period, of 147 subjects recruited who had no detectable parasites by microscopy on Giemsa-stained thick and thin blood smears and who met the inclusion criteria, 143 had P. falciparum infections detectable by PCR with msp1-specific primers. Some characteristics of these subjects including sex, haemoglobin type, mean PCV and their use of bed and/or window nets are given in Table 1. Fifty nine percent of recruited individuals were found to be anaemic, particularly amongst younger subjects: 63% of less than 1 year-olds, 33% of 1 to 5 year-olds and 17% of 6 to 15 year-olds had PCVs below the WHO threshold, respectively. This decrease in anaemia with age was significant (p < 0.0001, r = 0.56).

Antibodies reacting with rMSP1₁₉ were detected in 85% of the individuals with sub-microscopic infections. There was a significant increase in the proportion of responders to rMSP1₁₉ protein (p < 0.0001) and in the mean log titre of anti-MSP1₁₉ antibodies (Spearman correlation, p < 0.0001) with age, as shown in Table 2.

There was marked heterogeneity in the ability of polyclonal antibodies to compete with the MSP1₁₉-specific mAbs. The samples were divided into two groups: those containing highly competitive antibodies resulting in low binding of mAb to rMSP1₁₉ (≥ 50% inhibition of binding), and those containing poorly competitive antibodies resulting in high binding of mAb to rMSP1₁₉ (<50% inhibition of binding). As shown in Table 2, 39% (57 of 147) of samples contained antibodies that were highly competitive to mAb 12.10 binding to rMSP1₁₉. The 1- to 5 year-old age group contained the lowest prevalence of antibodies competitive with mAb 12.10 and although the inhibition of mAb 12.10 binding varied between age groups the difference was not significant. The inhibition of mAb 12.10 binding correlated strongly with the prevalence of total anti-MSP1₁₉ antibodies (Spearman correlation test, p < 0.0001) and with the mean titre of anti-MSP1₁₉ antibodies (Spearman correlation test, p < 0.001) (Table 3). Forty-six of the 147 samples (31%) contained antibodies that competed with mAb 12.8, and there was a correlation between the presence of such antibodies and age (p < 0.001). There was also a strong correlation of competition with mAb 12.8 and the prevalence and mean titre of polyclonal anti-MSP1₁₉ antibodies (p < 0.0001) (Tables 2 and 3). Overall, plasma samples containing antibodies highly competitive with both mAbs were also those with the highest titres of total anti-MSP1₁₉ antibodies (Table 3).

It was of interest to determine whether or not multiplicity of infection (MOI), and the numbers of different MSP1 types affected either the total amount of MSP1₁₉-specific antibody or of antibody that competed with the mAbs, and therefore the number of msp1 sequence block 2 types in the samples was used as a marker for MOI and MSP1 polymorphism in each individual. Genotyping of the polymorphic sequence block 2 revealed that RO33 was the predominant allele with 57% of the total, with the K1 and MAD20 allelic families representing 26% and 17%, respectively. Based on size polymorphism, only one RO33 allele was detected whereas the MAD20 and K1 families contained 8 and 13 distinct alleles, respectively (Figure 1). There was no preferential distribution of allelic families between age-groups. The mean minimum MOI based on the number of msp1 genotypes detected for individuals in each age-group is shown in Table 2. The highest MOI was found in the 6 to 15 years age group. Seventy-three of 147 (51%) isolates were found to have more than one P. falciparum genotype. There was no significant influence of age on the proportion of individuals with multiple versus single infections (denoted by M and S, respectively in Table 2) or on MOI. Overall, no correlation between MOI and the mean log titre of anti-MSP1₁₉ antibodies was detected. Similarly the MOI was not correlated with inhibition of mAb 12.10 binding. However, comparing the two groups M (multiple infection) and S (single infection), there was a higher prevalence of antibodies that competed with mAb 12.10 in samples from group M (Table 3, p = 0.04).

The proportion of individuals with anaemia decreased with age (Figure 2). Using logistic regression, the association of anaemia and the carriage of multiple P. falciparum genotypes on the total MSP1₁₉-specific antibody and the competitive activity with mAbs 12.10 and 12.8 were examined. Anaemia was significantly associated with total anti-MSP1₁₉ antibody (p = 0.0016) with a low prevalence of antibodies that competed with mAb 12.10; whereas harbouring more than 1 parasite genotype (group M) was associated (p = 0.04) with a greater prevalence of antibodies that competed with mAb 12.10. This was not observed with MAb 12.8: the presence of antibody competitive with mAb 12.8 was influenced neither by the presence of multiple P. falciparum genotypes nor by anaemia.