Antibody responses to malaria in people living in endemic regions are complex and represent the summation of multiple infections, typically beginning in infancy. Repeated infections are required to stimulate the non-sterile immunity found in adults living in malaria hyperendemic areas. Little is known about the immune responses to a single P. falciparum infection. In the group of previously healthy, mostly P. falciparum naïve, adults studied here, antibodies were measured to seven different malaria proteins including multiple MSP1 and MSP2 regions and alleles. Relatively short-lived antibody responses were found most commonly with MSP4/5/7 and MSP1 block 2 antibodies being significantly less frequent than the rest of the antibodies tested. In nine subjects, functional, invasion inhibiting MSP119 antibodies were shown to be present. Importantly, frequent antibody reactions to conserved regions of MSP2 have been documented for the first time in humans exposed to a single P. falciparum infection. Similarly, the MSP2 isotype response was not skewed towards IgG3 antibodies as previously shown . These features of the immune response to P. falciparum seen in naïve adults contrast markedly with the MSP2 repeat region dominated response found in malaria endemic populations .
MSP119 appeared to be particularly immunogenic in this population as shown by the presence of invasion inhibiting antibodies in 9/20 patients that were present one month after infection but mostly disappeared by six months. Of interest, a significant proportion (45%) of sera inhibited schizont rupture in growth assays, a property much more rarely encountered in sera from endemic areas (data not shown). Dilution of these sera revealed an underlying capacity to inhibit merozoite invasion. Functional MSP119 antibodies, as detected by this method, were correlated with protection from P. falciparum infection in Kenyan children and adults while MSP119 antibodies measured by ELISA were not . Other studies do not show the same level of correlation and the exact protective importance of this antibody subpopulation is not yet defined in all populations [21, 22]. Nevertheless, it seems reasonable to suggest that the presence of such antibodies indicates a capacity of convalescent sera to exert a level of anti-parasitic action after a single infection. It will be particularly interesting to discover whether similar levels of inhibitory antibodies are induced in children resident in endemic areas, after their first bout of malaria. This may represent a previously unrecognized mechanism for short-term host protection.
MSP1 block 2, by contrast, was much less antigenic. Only 4/20 of the adult population were positive for any MSP1 block 2 antibodies. This contrasts with previous field studies that also measured MSP1 block 2 antibodies to allelic family specific, short peptides. An age-dependant increase in antibodies to 58.3% of 10–15 year old Ghanaian children was found . The low frequency of antibodies MSP1 block 2 in this study does not appear to be due to failure to detect short-lived antibodies  as the patients were assayed immediately after their clinical episode of malaria with fine epitope mapping using 15 mer peptides that cover the MSP1 block 2 locus with significant redundancy. The epitopes recognized in this study were all from relatively constant regions of the MSP1 block 2 alleles. One patient who had no past history of malaria infection had antibodies detected to both K1 and RO33 allelic family antigens despite being infected with only a K1 family isolate. The observed low antigenicity of MSP1 block 2 in adults mostly after their first P. falciparum infection suggests that the potentially protective antibodies to this region  reflect a high frequency of previous episodes of malaria. The other antigens assayed with relatively low levels of seroreactivity were MSP4/5 and MSP7, which, like MSP1 block 2, appear to have higher rates of seropositivity in malaria endemic populations [9, 17].
For the first time, in this study, there is evidence of commonly occurring antibody responses to the conserved regions of MSP2. This contrasts with prevailing evidence that suggests the immunodominance of the central amino acid repeat region following repeated infections of humans  or immunization studies . Antibody responses to the conserved regions of MSP2 were documented in 15/20 of study patients. By contrast, antibodies from pooled PNG adult serum to this region were below the cut-off level so this result is highly significant. When the same methodology was used to study semi-immune Vietnamese teenagers and adults, only 1/15 had antibodies to the MSP2 conserved region . No measurable antibodies to the conserved regions of MSP2 were induced by vaccination with the 3D7 allele in five to nine year old children in the Combination B trial . A minority of non-immune Swiss travellers have been shown to have antibodies to the conserved regions of MSP2 by non-quantitative immunoblot .
Additionally, significant correlations between MSP2 allelic family constructs have been shown. More significant correlations between antibodies were found in individuals with a past history of one to two episodes of malaria. Even in previously naïve individuals, apparently cross-allelic family reactive antibodies were present. It appears that these are attributable to antibodies to the conserved region of MSP2. The correlations we have calculated are based on small numbers of samples. Previous studies of anti-MSP2 immune responses in populations living in hyperendemic areas have shown that antibodies are dimorphic family specific , but that there is extensive intra-dimorphic family immune response cross-reactivity . Vaccine-derived MSP2 antibodies were restricted to the 3D7 specific regions of the vaccine antigen . Limited inter-dimorphic family antibody cross reactivity has been recognized previously for MSP2 after patent malaria in two non-immune patients  and in this study, inter allelic family specific antibodies to MSP1 block 2 were found in one patient.
Protective antibodies to repeat regions of MSP2 have been shown to change from IgG1 to IgG3 dominated responses with aging indicating a maturation of the immune response [28, 29]. The patients studied here do not show this same isotype pattern and MSP2 antibodies were predominantly IgG1. Thus, these non-immune adults studied herein have MSP2 responses that differ in target and isotype from hyperimmune sera. It may be that the repeat region dominated MSP2 responses seen in malaria endemic regions result from the repeated infections experienced by infants with immature immune systems that do not recognise the conserved regions of MSP2. Similarly, the IgG3 dominated anti-MSP7 response in immune adults  was not seen in this study population. Previously naïve patients were also not found to have higher IgM responses to any of the antigens tested than adults with a past history of malaria.
Antibody responses trended to declining levels over the 6 months of follow up. At the final bleed though, antibodies to all antigens tested remained positive. The level of parasitaemia, not unexpectedly, had essentially no influence on the level of antibodies measured. In field settings with seasonal malaria transmission, MSP119 and MSP1 block 2 antibodies have been shown to decrease over dry seasons . Antibodies to chondroitin sulphate A-binding variant surface antigens have also been shown to be short-lived in women with pregnancy associated malaria but their presence correlates with protection against infection of the placenta . Temporary humoral immune responses seem to be common for malaria antigens.