A detailed description of the study area has been previously reported 161819. The study was carried out before the rainy seasons between 2004 and 2006 in the Daraweesh village, Gedaref State in eastern Sudan. Daraweesh is 450 km from Khartoum and 16 km from Gedaref town. It is inhabited by approximately 420 Arabic speakers of Fulani ethnic origin whose ancestors settled in this area about a century ago 20. The village economy is based on agriculture. Malaria transmission is markedly seasonal and unstable and annual peak parasite prevalence ranges from 1 to 40% in different years. The transmission is hypo-endemic and the acquisition of clinical immunity with age is not as obvious as seen in holo-endemic areas and thus malaria remains a problem in all age groups 20. Plasmodium falciparum is responsible for >96% malaria cases, the remainder being Plasmodium vivax and Plasmodium malariae. Anopheles arabiensis is the sole vector.

This study is a part of a longitudinal study of infection and immunity to malaria in Daraweesh, which has been going on since 1990. The study received approval from the ethical Committee of the University of Khartoum and national clearance from the Sudanese Ministry of Health. In this Daraweesh population, who are of Fulani origin, a study showed distinctly variable level of disease susceptibility 16. In the present study, data and samples from a cohort of 250 permanent residents (162 female, 88 male; median of age in 2005, 15.5 years, with the range 4–75 years), monitored clinically and parasitologically by passive case detection for up to 11 years were used. During the malaria transmission season, all individuals feeling unwell reported to the health team that was present in the village on daily basis. A total of 101 subjects from the sympatric non-Fulani ethnic neighbours – Masaleit (51%), Hausa (27.7%) and Four (21.8%) 14, were consecutively selected for the study (60 female, 41 male; median of age in 2005, 17.0 years, with the range 2–55 years). All blood samples were collected before the rainy season and all individuals were asymptomatic and negative for Plasmodium parasites.

Three ml of peripheral blood were collected from the individuals into vacuum EDTA tubes. The collected blood samples were centrifuged for 15 minutes at 250 g. The layer of white cells on top of red blood cells were collected into sterile cryotubes and stored frozen at -20°C for DNA extraction and FcγRIIa genotyping. The plasma was transferred into cryotubes stored at -20°C until use for antibody detection.

Genomic DNA was purified from the buffy-coat cells using a modified version of the Chelex-100 method and then stored at -20°C 21. In brief, 25 ml from the cells were incubated overnight at 4°C in 1 ml of 0.5% saponin in 1× phosphate-buffered saline (PBS). The pellet was washed for 30 min with PBS at 4°C, and the supernatant discarded. The pellet was boiled in 120 ml of 5% Chelex-100 in water for 15 min, and the DNA was collected in the supernatants after centrifugation at 250 g for 3 min.

The FcγRIIa genotype was determined using a modified version of a polymerase chain reaction (PCR) method 22. The PCR conditions were as follows: one cycle at 96°C for five minutes, 30 cycles at 94°C for 30 sec and 56°C for 30 sec, and one cycle at 72°C for 45 sec. The PCR product was digested for 2 h at 37°C using the restriction enzyme BstUI (Fermentas Inc. MD, USA), according to the manufacturer's recommendation. The enzyme cuts at one site in the R131 allele and twice in the H131 allele. The fragments were resolved by electrophoresis in a 2% agarose gel.

The levels of serum antibodies (IgG total and subclasses) to four malaria antigens (AMA-1, MSP 2 – 3D7 & FC27, Pf332-C231) were measured using enzyme-linked immunosorbent assays (ELISA), mainly as previously described 23. Briefly, EIA/RIA plates (Costar, MA, USA) were coated with AMA-1 at 1 μg/ml, 3D7-MSP 2 and FC27-MSP 2 at 1 μg/ml, and Pf332-C231 at 5 μg/ml. The AMA-1 protein, which has an N-terminal hexa-His tag, was expressed in Escherichia coli and refolded in vitro 24. Both the 3D7 and FC27 forms of MSP 2 were expressed in E. coli with C-terminal hexa-His tags. The expression and purification of these proteins will be described elsewhere (Hobba et al, manuscript in preparation). The recombinant Pf332-C231 corresponds to a 231-amino acid fragment in the C-terminal part of Pf332 25. The plates were incubated overnight at 4°C, and then blocked for 2 hrs with 0.5% bovine serum albumin (BSA) diluted in carbonate buffer (pH 9.6). Plasma samples diluted in incubation buffer (PBS + 0.5% BSA), 1:1,000 (IgG) and 1:400 (IgG1–4), were added in duplicate and incubated for 1 h at 37°C. The plates were than washed four times, and bound IgG antibodies were detected with goat anti-human IgG-ALP (1:2000) (Mabtech, Nacka, Sweden). IgG subclasses were analysed with their respective biotin conjugated mouse anti-human subclass specific monoclonal antibodies: mouse anti-human IgG1 1:1,000 (M15015, Clone NL16, SkyBio, Bedfordshire, UK), mouse anti-human IgG2 1:3,000 (555874, Pharmingen, Erembodegem, Belgium), mouse anti-human IgG3 1:1,000 (MH 1532, Caltag laboratories, Paisley, UK) and mouse anti-human IgG4 1:2,000 (B3648, Sigma, St. Louis, USA). Alkaline phosphatase (ALP) conjugated streptavidin (Mabtech) diluted 1:2,000 was added to detect bound antibodies of IgG2–4, while ALP-conjugated to goat anti-mouse Ig (Dakopatts, Glostrup, Denmark; 1:1,000) were used for IgG1 antibodies and the plates were developed with nitrophenyl phosphate (Sigma-Aldrich Chemie GmbH, Steinheim, Germany). The absorbance was read at 405 nm using a Vmax™ Kinetic microplate reader (Molecular devices, Menlo Park, USA) and the antibody concentrations were deduced from the log-log correlative coefficient of each IgG subclass standard curve (six dilutions of myeloma proteins of IgG1–4 subclasses), ranging from 0.01 to 3 μg/ml for IgG1, 0.001 to 0.3 μg/ml for IgG2, 0.001 to 0.1 μg/ml for IgG3 and 0.01 to 1 μg/ml for IgG4 according to the manufacturer's recommendation (Biogenesis, Poole, England).

FcγRIIa genotype, allele frequencies and antibody levels were analysed using SPSS version 10.0 (SPSS, Inc, Chicago, IL, USA). Logistic regression analysis was performed with age, sex and antibodies modelled as binary dummy variables (ranked into third distribution using the first third as the reference indicator). Associations were quantified using odds ratios [OR] with 95% confidence intervals [CI] that did not cross 1.00 with P value < 0.05, defined as statistically significant. The FcγRIIa-His/Arg131 group was used as a reference in the analyses, because this genotype is the most prevalent in human populations 26. An overall comparison of allele frequency using a 2 × 2 chi-square test was performed using the same software.

All individuals of the two study groups were genotyped for the FcγRIIa-H/R131 SNP. The H/H131 genotype was at higher frequency in the Fulani than in the non-Fulani ethnic group (36.0% for Fulani versus 17.8% for non-Fulani; unadjusted OR 2.67, 95% CI 1.48–4.83, P < 0.001 and adjusted OR 3.10, 95% CI 1.61–5.97, P < 0.001) (Tables 1 and 2). In contrast, there were no differences between the two groups in the frequency of R/R 131 genotype (17.6% for Fulani versus 20.8% non-Fulani; adjusted OR 0.99, 95% CI 0.52–1.89, P value 0.98) (Tables 1 and 2). However, when using the H/H131 genotype as a reference group in the statistical analysis, the heterozygote H/R131 genotype was found to be statistically significantly at higher frequency in the non-Fulani than in Fulani (46.4% for Fulani versus 61.4% for non-Fulani; adjusted OR 0.34, 95% CI 0.18–0.62, P < 0.001) (Table 2). A similar significant difference was seen for the R/R131 genotype, also being at higher frequency in non-Fulani as compared with Fulani (17.6% for Fulani versus 20.8% for non-Fulani; adjusted OR 0.36, 95% CI 0.17–0.77, P < 0.001) (Table 2).

A genetic difference between the two ethnic groups was also evident when comparing the FcγRIIa-R/H131 allele frequencies, the H131 allele being more common in the Fulani (0.60 for Fulani versus 0.49 non-Fulani; OR 1.54, 95% CI 1.09–2.17, P value 0.01) (Table 1). No such interethnic difference was seen for the R131 allele. The overall genotypic frequencies did not differ statistically among the different age groups within the ethnic groups, neither in the Fulani (P value 0.26) nor in the non-Fulani (P value 0.59). No gender dependent differences in genotype frequencies were seen (P = 0.61).

In this study, antibody responses to four P. falciparum blood-stage antigens, AMA-1, MSP2 – 3D7, – FC27 and Pf332-C231, were analysed in the Fulani and the non-Fulani ethnic groups. There were no statistically significant differences in anti-malarial total IgG antibody levels between the ethnic groups, and therefore the following analyses were limited only to the IgG subclasses. In general, IgG1 and IgG3 subclass antibodies were found to be at higher levels than IgG2 and IgG4 antibodies in both ethnic groups. However, the antibody levels varied for the different antigens. When comparing the antibody levels between the two ethnic groups, the non-Fulani showed significantly higher levels of IgG1 antibodies reactive with AMA-1 (P < 0.001), and also tended to have higher IgG3 antibody levels (P = 0.062) (Table 3). The IgG2 antibody responses reactive with AMA-1 were, however, statistically significantly higher in the Fulani than in the non-Fulani ethnic group (P = 0.024) (Table 3). While antibodies reactive with MSP2-3D7 showed a similar pattern of interethnic differences in IgG subclass levels as those reactive with AMA-1, IgG1 antibodies reactive with MSP2-FC27 were at higher levels in the Fulani as compared to the non-Fulani (P = 0.041) (Table 3). Regarding Pf332-C231 reactive antibodies, IgG3 and IgG4 subclasses were significantly higher in the non-Fulani than in the Fulani (P < 0.001) (Table 3). No differences in levels of antibodies of the different IgG subclasses were seen among the different ethnicities within the non-Fulani group.

In order to see if the FcγRIIa genotype affected the specific IgG subclass levels, using four different malarial antigens, the relationship between FcγRIIa genotype and IgG antibody subclasses was analysed. For the study population as a whole, including both the Fulani and the non-Fulani individuals, the H/H131 genotype was statistically significantly associated with higher levels of IgG2 (P < 0.001), IgG3 (P = 0.01) and IgG4 antibodies against AMA-1 (P = 0.01) (Table 4). Furthermore, there was a trend for an association of the R/R131 genotype with higher levels of IgG1 antibodies against AMA-1 (P = 0.05) (Table 4). Interethnic comparison of AMA-1 reactive antibodies in relation to FcγRIIa genotypes showed that non-Fulani H/H 131 carriers had significantly higher levels of IgG1 (adjusted OR 0.04, 95% CI 0.02–0.12; P < 0.001) and IgG3 (adjusted OR 0.31, 95% CI 0.17–0.57; P < 0.001) (Table 5). No such interethnic differences were found for the IgG2 and IgG4 antibody subclasses.

Regarding antibodies reactive with MSP2-3D7 or -FC27 in the study population as a whole, a similar pattern as with AMA-1 reactive antibodies was seen. The H/H 131 genotype carriers showed statistically significantly higher levels of IgG2 (P < 0.001 and = 0.01, respectively), IgG3 (P = 0.01 for both antigens) and IgG4 (P = 0.01 and < 0.001, respectively), while the R/R 131 genotype carriers showed higher levels of IgG1 antibodies (P = 0.01 for both antigens) (Table 4).

Interethnic comparison of MSP2-3D7 reactive antibodies in relation to FcγRIIa genotypes showed that non-Fulani H/H 131 carriers had significantly higher levels of IgG1 (adjusted OR 0.10, 95% CI 0.03–0.20; P < 0.001) and IgG3 (adjusted OR 0.49, 95% CI 0.26–0.91; P value 0.03), while Fulani H/H 131 carriers showed higher levels of IgG2 (adjusted OR 2.43, 95% CI 1.32–4.47; P < 0.001) and IgG4 antibodies (adjusted OR 2.76, 95% CI 1.45–5.26; P < 0.001) (Table 5). In contrast, an interethnic comparison for MSP2-FC27 reactive antibodies, showed that the non-Fulani H/H 131 genotype carriers had significantly higher levels of IgG2 (adjusted OR 0.45, 95% CI 0.24–0.85; P = 0.01) and IgG3 (adjusted OR 0.13, 95% CI 0.10–0.26; P < 0.001) than the Fulani, while no significant interethnic differences were found regarding IgG1 and IgG4 antibody levels with any of the genotypes (Table 5).

Regarding antibodies reactive with Pf332-C231 in the study population as a whole, the H/H131 genotype was statistically significantly associated with higher levels of IgG2 (P = 0.01) and IgG3 (P < 0.001) (Table 4). Furthermore, the R/R131 genotype was statistically significantly associated with higher levels of IgG1 antibodies reactive with this antigen (P = 0.01) (Table 4).

Interethnic comparison of Pf332-C231 reactive antibodies in relation to FcγRIIa genotypes showed that non-Fulani H/H 131 carriers had significantly higher levels of IgG2 (adjusted OR 0.22, 95% CI 0.11–0.44; P < 0.001) and IgG4 (adjusted OR 0.10, 95% CI 0.03–0.20; P < 0.001), IgG3 antibodies were at significantly higher levels in H/H 131 carriers of the Fulani group (adjusted OR 4.25, 95% CI 2.28–7.91; P < 0.001) (Table 5).