Background
Malaria infections of humans, particularly that due to
Plasmodium falciparum continues to be a major cause of morbidity and mortality in tropical countries. There is an urgent need for the development of efficacious control measures, one component of which could be a safe, effective and affordable malaria vaccine against
P. falciparum. It is widely believed that any such vaccine will need to incorporate multiple antigens from the various stages of the parasite's complex life cycle [
1].
The surface of the asexual stage merozoite form of
P. falciparum is composed of a number of proteins that are the targets of immune attack by antibodies. One of these proteins is Merozoite Surface Protein 4 (MSP4), a relatively abundant glycosylphosphatidylinositol-anchored protein that contains a single epidermal growth factor (EGF)-like domain adjacent to the carboxyl terminus of the protein [
2,
3]. Although the function of MSP4 is not known, the
msp4 gene is refractory to genetic deletion and it is thus thought to be essential for parasite replication in
in vitro culture and presumably also in the human bloodstream [
4]. Several features of MSP4 make it an attractive vaccine candidate. Firstly, MSP4 is exposed on the merozoite surface making it available for antibody binding and anti-MSP4 antibodies are readily detected in people living in malaria endemic regions [
5,
6] suggesting a possible role for these antibodies in human immunity to malaria. Secondly, MSP4 shows a high degree of conservation among
P. falciparum isolates [
7‐
9] minimizing the possibility of immune evasion secondary to strain-specific antibody responses. Thirdly, immunization of mice with recombinant
Plasmodium yoelii MSP4/5, a homologue of both MSP4 and the related antigen MSP5, protects mice against lethal parasite challenge [
10,
11]. Protection is enhanced when MSP4/5 is immunized in combination with
P. yoelii MSP1
19[
12] suggesting that it would be an attractive addition to a multi-antigen vaccine containing MSP1
19.
A panel of nine anti-MSP4 monoclonal antibodies (Mabs) that recognize distinct epitopes of the antigen were produced and characterized. These antibodies were tested in a competition enzyme-linked immunosorbent assay (ELISA) against human immune sera collected from P. falciparum-infected subjects to analyse the binding characteristics of anti-MSP4 antibodies induced by natural infection. The ability of polyclonal and monoclonal anti-MSP4 antibodies to inhibit parasite growth in vitro were also assessed in this study.
Discussion
The natural immune response to malaria is highly complex involving both antibodies and cell mediated immunity [
21]. The clinical symptoms of malaria are caused by the asexual blood stage. Antibodies are particularly important in protective immunity during this stage as they can bind to antigens on free merozoites and inhibit erythrocyte invasion either directly or in association with cells to limit disease pathogenesis and clinical symptoms. Evidence for the protective role of antibodies in clinical malaria are provided from studies where passive transfer of antibodies from immune adults was able to successfully treat children with severe
P. falciparum infection [
22]. However, the finding that long-term exposure to the parasite resulting in multiple antibody specificities is necessary to generate protection from malaria indicates that the establishment of protective antibodies is a complex process. The breadth and magnitude of the antibody responses to multiple merozoite antigens are shown to be associated with protection from clinical malaria [
23]. The antibody response to merozoite surface protein 1 (MSP1) has been extensively studied and particular areas of the protein have been found to be immuno-dominant [
17]. For example, antibodies against the 19 kDa C-terminal fragment of MSP1 are the major component of the invasion inhibitory response in humans immune to malaria [
19,
24]. The protective effect of these anti-MSP1
19 antibodies is further dependent on their fine specificity, rather than mere prevalence or titre [
25]. Similar studies suggest the existence of immune-dominant functional domains that are the target of protective antibodies to apical membrane antigen-1 (AMA-1) [
26,
27]. In contrast, our previous immuno-epidemiological studies showing recognition along the length of MSP4 [
6], as well as the fine mapping and functional studies presented herein, suggest the possibility that acquisition of protective antibody immunity to MSP4 could be a cumulative process with a requirement for the generation of multiple antibody specificities.
In this study, nine monoclonal antibodies specific for recombinant MSP4 protein expressed in
Escherichia coli were produced and characterized. These Mabs recognize six distinct epitopes of MSP4, including two conformational epitopes in the C-terminal region containing the EGF-like domain. The remaining Mabs recognize epitopes that are not affected by reduction and alkylation and are presumably linear, with two in the C-terminal region of MSP4 and two in the central region of the protein. The reactivity of Mab H12 was interesting in that it reacted only with the reduced and alkylated form of full-length MSP4 but mapped to a region of the molecule, MSP4A, that has no disulphide bonds. This suggests that there are long range conformational effects which alter the shape of the N-terminus of MSP4 when disulphide bonds in the C-terminus are interrupted. This is in accord with previous observations in which the redox state of the EGF-domain was crucial for the antigenicity of the entire protein, including regions that are not immediately adjacent in the primary structure and did not contain disulphide domains such as the N-terminus [
5].
This panel of antibodies was used in competition ELISA to analyse the binding specificities of naturally acquired antibodies to MSP4 in individuals from a malaria-endemic region of Vietnam. The three tested regions of MSP4 were readily recognized by the sera examined. The existence of persistent heterogeneity was observed among individuals in the level of antibody reactivity and the spread of epitope recognition. The reactivity of anti-MSP4 antibodies in the sera tested was considerably higher for epitopes in the central region of MSP4, in particular the MSP4B region. This result is also in agreement with our previous findings [
6] that the percentage of positive responses was higher for MSP4B (92.5%) compared to the EGF-like domain containing MSP4D region (71.3%). This is also in accord with other reports where cysteine-rich regions of proteins have been shown to be less immunogenic [
28]. In contrast, six of the nine Mabs generated in this study by immunization of mice with rMSP4 were specific for the EGF-like domain.
MSP4 is a surface exposed protein that is highly conserved. The finding that antibodies target multiple epitopes spread across the entire length of the molecule may provide some insight into the mechanisms for co-evolution of parasite with its human host. It is possible to speculate that the sequence conservation of MSP4 maybe partly caused by the fact that immune selective pressure needs to target multiple epitopes simultaneously rather than individual epitopes. In addition to helping in understanding the nature of the antibody response to MSP4 during infection, the panel of monoclonal antibodies will also serve as valuable reagents for testing product integrity, purity, antigenicity and immunological activity during the highly complex and stringently regulated process of MSP4 vaccine development and manufacture.
It is shown that polyclonal rabbit antisera raised against the full length MSP4 protein are able to inhibit parasite growth
in vitro. In contrast, rabbit antisera specific for MSP4 fragments corresponding to three of the regions of MSP4 were unable to inhibit parasite growth. Similarly, the individual Mabs had no inhibitory effect on the growth of the parasite. These results are consistent with those previously obtained using an
in vivo P. yoelii murine malaria challenge model in our laboratory. Immunizations carried out with recombinant PyMSP4/5 were able to induce immune responses that protect mice against challenge with a lethal dose of
P. yoelii parasites [
10]. In contrast, immunizations with PyMSP4/5 fragments did not confer protection (L. Kedzierski, unpublished data). One explanation for this is that an antibody response directed to multiple epitopes spread across the molecule is necessary for arresting parasite growth. Since the combination of three Mabs directed to different regions of the MSP4 molecule were also unable to produce a growth inhibitory effect, it is likely that a polyclonal antibody response to more than three epitopes is necessary. Alternatively another possibility is that antibodies directed to epitopes other that those represented by the Mabs are required for invasion inhibition. These epitopes maybe ones that are only present in the complete folded sequence and accessible on the merozoite surface. The lack of inhibitory activity despite the high affinity binding of the Mabs may also be related to the function and size of the MSP4 molecule. In the case of AMA-1, inhibitory antibodies bind and block epitopes that are involved in the parasite invasion process. MSP4 is a small molecule (270 amino acids) and of relatively low abundance on the surface of the merozoite. Hence, the binding of antibodies to MSP4 alone may not be sufficient to cause high levels of parasite dysfunction or invasion inhibition, but MSP4-specific antibodies may contribute in combination with antibodies specific for other merozoite surface proteins.
Following schizont rupture and merozoite release, re-invasion occurs within seconds. Therefore, antibodies will only be useful if they are available at very high concentrations and/or are of high affinity. Consistent with this picture, the observations in this study showed that growth inhibitory capacity is directly proportional to antibody titre. Similar observations have been made in field studies in malaria endemic areas; under natural exposure, immunity to malaria results from high titre antibodies to multiple antigenic targets [
23]. These observations provide support for the production of combination blood stage vaccines.
The capacity to induce experimental antibodies with growth inhibitory activity (GIA)
in vitro is currently being used as a criterion for selection of blood stage antigens that are to be incorporated into a subunit malaria vaccine. Growth inhibition
in vitro is considered a functional assay because it measures the capacity of antibodies to bind and inhibit parasite invasion and/or growth. AMA-1 [
27] and MSP1 [
29] are the blood stage antigens currently in the most advanced stages of being developed as human malaria vaccines. Antibodies to AMA-1 so far have been able to show consistently high levels (> 80%) of growth inhibition
in vitro[
30,
31] and much attention has focused on this protein as a promising vaccine candidate against malaria. In contrast, anti-MSP1
19 antibodies show more moderate inhibitory activity [
29,
32,
33]. If growth inhibitory activity is sufficient for protection, then an effective malaria blood-stage vaccine would likely need to be made up of multiple antigens. Growth inhibition leading to decreased replication through the erythrocytic cycle is likely to be a result of the collective immunogenic ability of the multiple antigens. On this basis a growth inhibitory capacity of 40% observed for MSP4 as a single antigen would be a significant contribution. Further studies are required to determine if the level of growth inhibition is increased when antibodies directed to several merozoite surface antigens are tested in combination.
Despite extensive work the value of using GIA as the sole correlate of immune protection from
Plasmodium infections is still unclear. The validity of this
in vitro assay may be confirmed once vaccine trials are carried out in humans and
in vitro inhibition is shown to correlate with
in vivo protection. A recent Phase IIa trial of AMA-1 has shown disappointing results in this regard [
34]. Antibodies classically act in conjunction with other mechanisms of immunity, such as complement mediated destruction of the pathogen and enhanced phagocytic activity mediated by antibody opsonization. These mechanisms are also likely to be operating in immunity to blood stage malaria. Interestingly, the majority of human antibodies recognizing MSP4 were of the IgG3 and IgG1 isotypes suggesting that they may play a role in opsonization and complement-mediated destruction of free merozoites [
6].
The Antibody Dependent Cellular Inhibition or Cytotoxicity assay (ADCI or ADCC) is another
in vitro assay that can be used to determine antibody efficacy in co-operation with monocytes to limit
P. falciparum growth [
35,
36]. Interestingly, a positive ADCI has been demonstrated for other blood stage antigens MSP3 [
37] and GLURP [
38], despite the lack of
in vitro growth inhibition by antibodies alone. In view of this finding, our antisera raised against MSP4 are currently being tested for ADCI activity.
Apart from antibody-mediated immunity, cellular mechanisms such as the production of IFNγ-producing T cells are also important in mediating protection against blood stage malaria [
39,
40]. T cells that recognize dominant epitopes in vaccine proteins are essential for long-term protection against malaria in animal models. Preliminary work was carried out to predict the possible T cell epitopes of MSP4 in humans that bind across diverse HLA molecules using an epitope prediction algorithm [
41]. All four regions of the molecule contained areas that were predicted to have T cell epitopes and these areas also overlapped with the three peptides containing B cell epitopes. These findings have important implications for vaccine design enabling incorporation of peptide antigens that are able to stimulate strong T and B cell responses to MSP4.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
RLC conceived the study. HdeS and SS designed and performed the experiments and analysed the data. HdeS wrote the manuscript. MP and SK helped with revision of manuscript. LW, CB and MP contributed information and reagents. All authors read and approved the final manuscript.