Acquisition of antibodies to merozoite surface protein 3 among residents of Korogwe, north eastern Tanzania

This study was conducted in Korogwe district, north-eastern Tanzania. The district is about 100 kilometers inland from Tanga. The population of Korogwe district in year 2002 (National census survey) was estimated to be 261,004, with a growth rate of 1.4% per annum. The area is topographically stratified into three strata namely; lowland, intermediate and highland. The strata are characterized by marked differences in malaria transmission profiles. The altitude of Korogwe district ranges from 300 - 1,200 meters above sea level (mASL). In these areas, malaria transmission decreases with increasing altitude [17, 18] and is highest during and following the long rainy season, which usually extends from March through July. Low transmission is experienced during short rains of October - December. However, a recent study has shown non-significant differences in malaria prevalence between the two seasons [19].

In Korogwe district, the estimated average entomological inoculation rate (EIR) ranges in the past from 30-100 infective bites per person per year [17]. In lowland villages of Korogwe, malaria is perennial with peak seasons during and just after the rain [20]. P. falciparum is the predominant malaria species accounting for a little over 90% of all infections, the rest being P. malariae and P. ovale [21]. In Korogwe District Hospital, malaria is the leading cause of admission and deaths among underfive children [22].

Study villages were grouped into three strata namely; lowland, intermediate and highland based on the altitude and level of transmission. Mng'aza village was categorized as lowland stratum, Kwashemshi village as intermediate while Magundi, Kwamhanya and Vugiri villages were grouped into highland stratum. The study involved individuals aged 0-19 years. The details of the study villages are as explained elsewhere [19].

This study was part of a longitudinal project being conducted in Korogwe. Samples used in this study were obtained from a cross sectional malaria survey conducted in May and June 2006 [19]. Plasma samples from 492 individuals (66 lowlands, 121 intermediate and 305 highlands) were analyzed by Enzyme Linked ImmunoSorbent Assays (ELISAs). A sample size of 89 samples from each stratum was estimated assuming a proportional mean difference of antibody levels to MSP3 between any two strata to be 30% and coefficient of variation of 90%; this variation has been also shown to vary with transmission elsewhere [23]. Thus, a minimum sample size of 270 was required from the three strata. For the purpose of assessing the effect of age on acquisition of antibodies, individuals were categorized into four age-groups 0-2, 3-4, 5-9 and 10-19 years. Over sampling was done in the highland area, because the villages were initially earmarked for MSP3 phase 1b vaccine trial [24].

Ethical clearance was granted by the Medical Research Co-ordinating Committee of the National Institute for Medical Research and the Ministry of Health and Social Welfare, Tanzania. Prior to commencement of the study, sensitization meetings were held in each village to explain the objectives and methodology of the study as well as seeking for community consent. Informed consent was obtained in writing from all individual participants or parents/guardians in case of children during the survey as described elsewhere [19].

Clinical examination was done in all 492 individuals at a central point in each village. Venous blood (about 3 ml) from each individual was drawn into vacutainer containing EDTA anticoagulant. Out of the 3 ml of blood collected, 8 μL was used for preparation of thick and thin blood smears for malaria parasite diagnosis, species identification and quantification. Also, 10 μL of blood of each individual was used for estimation of haemoglobin using Haemocue^® machine. The remaining blood was centrifuged at 2,000 revolutions per minute (rpm) to obtain plasma, packed red blood cells (RBCs) and buffy coat. Plasma was stored at -80°C and later analyzed for immunological assays.

Plasma samples were used to determine levels of IgM, total IgG and IgG subclasses (IgG1 & IgG3) responses to MSP3 malaria vaccine candidate by Enzyme Linked ImmunoSorbent Assay (ELISA) technique as described elsewhere http://​www.​amanet-trust.​org. Briefly, wells of the microtiter plates (Nunc, Roskilde, Denmark) were coated with 100 μL at 0.5 μg/ml of a recombinant protein covering the nonpolymorphic carboxy-terminal region of the MSP3 protein (aa 212 - 380) [10]. The antigen was supplied as a stock solution of 2.4 mg/ml and had to be diluted so as to obtain a working concentration of 0.5 μg/ml. The plates were incubated overnight at 4°C, washed four times and blocked using (150 μL per well) 3% skimmed milk in phosphate buffered saline (PBS) -Tween 20 (Blocking Buffer) for 1 hour. Plasma samples (100 μL/well) at dilutions of 1:200 (for total IgG and IgM) and 1:50 (for IgG subclasses) were added in duplicate wells and incubated at room temperature (20 - 24°C) for 2 hours. Plates for IgG1 and IgG3 were then incubated for 1 hour with monoclonal mouse anti-human IgG1 and IgG3 (Skybio, Wyboston Bedfordshire, UK) as secondary antibodies at dilutions of 1:2000, and 1:5000 respectively. The plates were washed four times between aforementioned steps. Colour development was achieved either with horseradish peroxidase-conjugated goat anti-human total IgG and IgM (Caltag, Burlingame, CA) respectively, or horseradish peroxidase-conjugated goat anti-mouse IgG (Caltag, Burlingame, CA) for IgG subclasses followed by H₂O₂ with tetramethylbenzidine (TMB) substrate (Kem-En-Tec diagnostics, Kuldyssen, DK). Absorbance was read at 450 nm with reference at 650 nm. For each antibody isotype, a calibration (standard) curve was obtained from two-fold serially diluted purified myeloma proteins coated directly to duplicate wells of the first two columns of each plate. First well concentrations in ng/ml were: 100 (IgG & IgG3), 500 (IgG1) and 250 (IgM). Four pooled plasma from clinically immune adult Liberians and two Danish donors never exposed to malaria were used as positive and negative controls respectively. Antibody units were calculated using the Afro Immuno Assay (AIA) Network protocol of 2005 volume 1.01 excel based curve fitting program as described elsewhere [25].

All data were entered and verified in Microsoft Access while statistical analysis was done using R version 2.8.0 and STATA software version 8 (stata Corp., College station, TX). All antibody levels were transformed to normality using the log transformation. Malaria parasite densities were converted from parasites per 200 white blood cells (WBCs) to parasite per microliter (μL) by multiplying by 40. A multiple regression analysis was used to assess variables associated with antibody levels. A p-value < 0.05 was considered significant.

Table 1 gives the baseline characteristics of individuals living in the three strata during May/June 2006 cross-sectional survey. Out of the 492 selected individuals, 305 (62%) were from highland stratum, 121 (24.6%) were from intermediate stratum and 66 (13.4%) were from lowland stratum. Overall, malaria parasites prevalence was 20.5% (101 out of 492). Distribution of P. falciparum parasite prevalence in the three strata were 50.0% (95%CI: 37.4 - 62.6) in the lowland stratum, 23.1% (95%CI: 16.0 - 31.7) in the intermediate stratum and 9.8% (95%CI: 6.7 - 13.7) in the highland stratum. Intermediate stratum had low malaria parasite prevalence but its geometric mean parasite density (GMD) among the positive cases was the highest among the three strata (424 rings/μL; 95%CI: 187 - 964). Bednet use was highest (91.4%) in the intermediate stratum. Spleen rate was higher in the lowland (30.8%) compared to the other two strata.

Individuals living in lowlands had higher IgG1 antibody levels compared to negative controls while individuals aged 10-19 years had higher mean levels than positive control (Figure 1). Results show that levels of IgG1 adjusted for other covariates were significantly lower in both intermediate by 1.653 (95%CI: 1.218 - 2.088; p < 0.001) and in highland strata by 1.229 (95%CI: 0.853 - 1.605; p < 0.001), when compared to the lowland stratum. Increase in mean haemoglobin concentration (Hb) by 1 g/dL was associated with decrease in IgG1 levels by 0.100 (95%CI: 0.008 - 0.192; p = 0.035). No statistically significant association was found between IgG1 levels and presence of P. falciparum parasites (Table 2).

Individuals living in lowlands had higher IgG3 antibody levels compared to negative controls while individuals aged 10-19 years had higher mean levels than positive controls (Figure 1). A pattern similar to that of IgG1 antibody levels was seen for IgG3, with intermediate and highland strata having significantly lower mean antibody levels compared to lowland stratum (Table 2).

Individuals living in highlands had higher total IgG antibody levels compared to negative controls while individuals aged 10-19 years had higher mean levels than positive controls (Figure 1). Mean levels of total IgG were higher in highland stratum compared to intermediate and lowland strata (Figure 1). Levels of total IgG in log scale as shown in table 2, were significantly lower in the intermediate stratum by 1.168 (95%CI: 0.723 - 1.613; p < 0.001) and higher in the highland stratum by 0.742 (95%CI: 0.358 - 1.126; p < 0.001) compared to the lowland stratum. Individuals who had P. falciparum infection had significantly high levels of total IgG by log 2.552 (95%CI: 1.045 - 4.059, p < 0.001). Adjusting by strata and age among individuals who had positive blood smears, IgG levels were associated with parasite density, whereby, the log mean levels of IgG decrease was 0.227 (95%CI: 0.064 - 0.391; p = 0.007).

Individuals living in highland stratum had higher IgM antibody levels compared to negative controls while individuals aged 10-19 years had higher mean levels than positive control (Figure 1). Mean levels of IgM increased with increasing altitude, the highest levels being seen in the highland stratum (Figure 1). Log mean difference in IgM levels in the highland stratum was 2.997 (95%CI: 2.617 - 3.377; p < 0.001) and that of intermediate stratum was 2.141 (95%CI: 1.702 - 2.580; p < 0.001) when compared to the lowland stratum (Table 2). IgM levels were also associated with bednet use, where in individuals who mentioned to use bednets, log mean levels were lower by 0.338 (0.062 - 0.614; p = 0.017). Adjusting by strata and age among individuals who had positive blood smears, IgM levels were associated with parasite density, whereby, the log mean levels of IgM decrease was 0.165 (95%CI: 0.044 - 0.286; p = 0.008). Individuals using bednets had significantly (p = 0.017) lower IgM levels (Table 2).