Cross-Species Immunity in Malaria Vaccine Development: Two, Three, or Even Four for the Price of One?

For millennia, malaria has been one of the most deadly human infectious diseases in the world, and it is still the cause of about 2 million deaths per year, mainly among young children and pregnant women. Malaria is caused by protozoan parasites of the genus Plasmodium, four species of which are infectious for humans: Plasmodium falciparum (the most deadly species, responsible for 80% of infections worldwide and 90% of malaria-related deaths), Plasmodium vivax, Plasmodium malariae, and Plasmodium ovale (56).

The “gold standard” for full and sterile immunity against malaria—immunization with Plasmodium whole-organism, radiation-attenuated sporozoites (RAS)—was described for mice, monkeys, and humans over three decades ago (12, 14, 45a). The original RAS immunization strategy, however, usually required either the bites of many irradiated mosquitoes on multiple occasions or intravenous inoculation of sporozoites, both of which were considered impractical for mass vaccination campaigns (36, 45a). The advent of recombinant DNA technology in the 1970s led many researchers to believe that the development of vaccines for the majority of known diseases, including malaria, was imminent (45). Although significant achievements in vaccinology have indeed been made using this technology (e.g., the pertussis vaccine [47]), the quest for a malaria vaccine remains unfulfilled. Several potential candidate vaccines have progressed to clinical trials recently (http://www.who.int/vaccine_research/documents/RainbowTable_ClinicalTrials_December2006.pdf), and many others have been the subject of preclinical assessments (24, 42; http://www.who.int/vaccine_research/documents/RainbowTablePreclinical_December2006.pdf). Disappointing results with either DNA- or subunit protein-based vaccines for malaria have led recently to a renewed effort focusing on scaled-up production of P. falciparum RAS and optimization of the immunization regimen for mass vaccination using this attenuated whole-organism-based approach (36; http://www.sanaria.com). In the latter context there have been other exciting recent developments. In rodent malaria models, both Plasmodium berghei genetically attenuated parasites (PbGAP) and Plasmodium yoelii GAP and P. yoelii sporozoites inoculated under cover of chloroquine (PyLUCS) have been shown to confer similar degrees of protection (2, 31, 44, 44b, 64). Together, these different types of whole-organism, live attenuated sporozoites constitute a powerful tool that should help to unravel the host mechanisms leading to protection, thus providing knowledge that may be crucial for development of the long-awaited vaccine. Intriguingly, however, the immunity mediated in mice by RAS and PbGAP (p36p⁻ type) is not plasmodial species specific but in both cases confers protection against challenge infection with heterologous rodent plasmodial species (19, 45b).

Four different Plasmodium species are infectious for humans (P. falciparum, P. vivax, P. malariae, and P. ovale). The first hint that sporozoites may induce cross-species immunity came from medical observations made in the middle of last century. At that time, inoculation of plasmodial parasites, either by injection of infected blood or by allowing sporozoite-infected mosquitoes to bite patients, was investigated as a treatment for neurosyphilis in Europe and the United States (11, 15, 27, 66). Reports of such treatments indicate that inoculation of P. malariae (either blood from infected donors or sporozoites), but not inoculation of P. vivax or P. ovale, resulted in reduction of parasitemia levels and the frequency of fever following subsequent infection with P. falciparum (15, 27). Based on these observations, Collins and Jeffery proposed that P. malariae might share common antigens with P. falciparum that elicit an immune response strong enough to affect the development of P. falciparum (15). Developing a vaccine that could induce simultaneous protection against P. falciparum and P. vivax or even more human plasmodial species would have enormous economic, safety, and manufacturing advantages. However, the issue of cross-species immunity in malaria, specifically in the context of live attenuated whole-organism-based protection, has been largely ignored in the last few decades. Here, we provide an overview of the available information concerning Plasmodium cross-species immunity and discuss the potential application of this information to the development of a vaccine for malaria.

CROSS-SPECIES IMMUNITY IN LIVER-STAGE MALARIA

Once injected into the mammalian host by an infected anopheline mosquito, plasmodial sporozoites make the journey via the blood vasculature to their target organ, the liver (1, 43, 44, 56). Once they are there, development proceeds within hepatocytes, with each sporozoite giving rise to a schizont that progressively matures until tens of thousands merozoites are formed (56). This intrahepatocytic schizont development is not associated with clinical symptoms.

Very little is known about the capacity of live attenuated human plasmodial sporozoites to induce cross-species immunity. In immunization studies with either P. vivax or P. falciparum RAS performed with human volunteers, no cross-species protection was reported (25). However, this conclusion relied on results for a single volunteer who was immunized six times via the bites of P. falciparum RAS-infected mosquitoes. This individual was then shown to be fully protected against challenge with intact P. falciparum sporozoites on three separate occasions, but not against a single challenge infection performed later with intact P. vivax sporozoites. Due to the obvious difficulties in conducting research tests with human volunteers, this type of experiment was not repeated. Given these conditions, it is therefore difficult to draw any firm conclusions concerning the existence of cross-species protection induced by P. falciparum and P. vivax RAS. Malarial sporozoites that infect other nonhuman mammalian hosts do, however, confer protection against challenge infection with sporozoites of a heterologous species. Mice inoculated with P. falciparum sporozoites, for example, were shown to be protected against subsequent challenge with P. berghei sporozoites (57). This is clear evidence that cross-species effects can be mediated by common malarial antigens; i.e., plasmodial proteins that share a high degree of homology can induce cross-reactive immune responses. The effects observed in the study cited above may have been due to sporozoite-specific antigens since P. falciparum sporozoites are unable to infect murine hepatocytes (54). Obviously, not all the antigens expressed by one species contribute to such a cross-species effect in different hosts. Sera of mice immunized with peptides comprising B-cell epitopes present in the central immunodominant region of circumsporozoite protein (CSP) of P. yoelii, an antigen that stimulates both humoral and cellular responses, were unable to inhibit P. falciparum intrahepatic development in human primary hepatocytes, for example (21).

The vast majority of our knowledge concerning cross-species immunity and immunization with live attenuated plasmodial sporozoites was derived from experiments performed with murine models. Almost 40 years ago, Nussenzweig and colleagues showed that P. berghei RAS-immunized mice were fully protected against challenge with either Plasmodium chabaudi or Plasmodium vinckei sporozoites (45b). Mice immunized with P. chabaudi RAS were also fully protected against challenge with P. berghei sporozoites (45c). The protection observed seemed to be partially dependent on a blood factor, most probably antibodies (45b). Our recent results showed that the immune response elicited by both P. berghei RAS and GAP (p36p⁻) sporozoites provided partial protection against infection with P. yoelii (19). When two different mouse strains that were previously immunized with live attenuated P. berghei parasites were challenged with different doses of P. yoelii sporozoites, there were significant reductions in the parasite burdens in the livers of 99.9 and 94.6% of RAS- and p36p⁻-immunized mice, respectively. Moreover, these immunized mice had longer prepatent periods (which increased from 0.7 to 3 days) and significantly lower peak parasitemias (which were more than 50% lower), and most remarkably, the majority of the RAS-immunized mice (but fewer p36p⁻-immunized mice) displayed complete sterile protection. The absence of full protection in all mice contrasts with the findings of Nussenzweig and colleagues, but the discrepancy may reflect the different immunization and challenge doses, protocols, and mouse strains used (19, 45b, 45c). To the best of our knowledge, no investigations of the capacity of other GAP to induce cross-species immunity have been performed yet. In mice immunized with live attenuated P. yoelii parasites (PyLUCS) and challenged with P. berghei, the levels of parasitemia were similar to those of the control mice (50; L. Rénia, submitted for publication). Nevertheless, a small delay in the onset of blood-stage parasitemia was observed, possibly indicating low-level cross-species protection. Thus, cross-species protection can indeed be demonstrated in different murine malaria models, but it appears to depend on the experimental conditions used.

MECHANISMS OF SPOROZOITE-INDUCED CROSS-SPECIES IMMUNITY

In the absence of relevant detailed information, one can only speculate about the mechanisms that mediate cross-species protection, based on observations of experimental malaria in mouse models. It seems quite plausible that in the case of liver stages, CD8⁺ T cells play a key role in this type of immunity. Together with other factors, these cells are essential for induction of RAS-, uis3⁻ PbGAP-, and PyLUCS-induced homologous protection (2, 3, 17, 18, 29, 44c) and most probably also other GAP-induced immunity, although the latter remains to be confirmed. A cytotoxic T-cell clone obtained from a mouse immunized with P. yoelii RAS recognized an epitope on both P. yoelii and P. berghei CSP (67). This clone conferred protection to naïve mice against a challenge infection with either P. berghei or P. yoelii sporozoites, supporting the belief that cross-species immunity elicited by RAS indeed relies on CD8⁺ T cells. In contrast, T-cell clones obtained from mice immunized with P. yoelii synthetic peptides corresponding to segments of CSP do not affect P. berghei sporozoite infection (49, 51, 53). However, immunization with whole-parasite RAS leads to a broader and more effective immune response than immunization with subunit vaccines, as discussed above. Most probably, the quantity and quality of T cells stimulated by RAS are enhanced, and therefore the cells are able to exert the species-transcending effect observed by Weiss and colleagues (67).

Apoptosis may play a prominent role in cross-species protection. It is known that in order to survive, plasmodial sporozoites prevent host cell apoptosis during their intrahepatocytic development (34, 63). We have shown that the capacity of p36p⁻ PbGAP to prevent infected hepatocyte apoptosis, in turn, is severely impaired; such cells are eliminated from the liver very early during parasite development (64). Host cell apoptosis has also been shown to occur during the early stages of RAS development (35), and we observed a similar but more widespread outcome during p36p⁻ PbGAP intrahepatocytic growth (64). Apoptotic infected host cells can potentially provide a huge array of intracellular pathogen-derived antigens to antigen-presenting cells (e.g., dendritic cells), as has been observed for Plasmodium, Mycobacterium tuberculosis, and Salmonella (35, 68). Thus, a plausible conclusion is that the heterologous protection against P. yoelii mediated by p36p⁻ PbGAP might be at least partially the result of apoptotic infected hepatocytes. The murine plasmodial species are considered to be phylogenetically distantly related (46); however, P. berghei has average protein identities of 88.2% with P. yoelii and 83.2% with P. chabaudi (23), suggesting that several antigens common to P. berghei and to the two other murine plasmodia could be presented to the immune system via such an apoptotic event. The cross-protective response elicited by one attenuated parasite species would therefore be strong enough to act against a different species. We speculate that higher doses of RAS or p36p⁻ PbGAP per immunization would increase the possibility of species-common antigen presentation, leading to complete sterile protection against P. yoelii and other murine plasmodia.

Antibodies are also crucial for RAS-induced immunity (55a) and, most probably, for live attenuated sporozoite-based cross-species immunity. Passive transfer of serum from P. berghei RAS-immunized mice was able to fully protect 11% of the mice challenged with P. vinckei (45a). The serum contained antibodies with specificity for P. berghei antigens that seemed to have considerable cross-reactivity with P. chabaudi and P. vinckei antigens (45a). Since RAS-induced homologous immunity is dependent on both cellular and antibody-based responses (17, 55a), one could predict that these responses must also be crucial for RAS-induced cross-species protection. The fact that the PyLUCS-mediated protective effect does not rely on antibodies (2) may then explain the poor capacity of this attenuation approach to confer cross-species protection. PbGAP uis3⁻-mediated protection, additionally, does not rely on antibodies (44c). Cross-species protection experiments using this GAP should definitively elucidate the role of antibodies in this type of immunity. The role of antibodies in uis4⁻, p36p⁻, and p36⁻/p36p⁻ GAP-based immunity (either homologous or heterologous protection) is currently unknown.

CROSS-SPECIES IMMUNITY TO ASEXUAL BLOOD-STAGE PARASITES

Cross-species interactions are also observed with plasmodial asexual blood stages, but these interactions are poorly understood and have been less well studied than the cross-species protection associated with exoerythrocytic stages discussed above. In fact, the existence of cross-species immunity is contested (9, 41, 52); for example, it is argued that parasites coexisting within the same host tend toward antigenic diversity in order to promote their own survival, leading to species-specific immunity. In this context, the low level of homology between some human Plasmodium species' surface proteins (e.g., CSP and merozoite surface protein-1) theoretically should be an obstacle to the establishment of cross-species protection. In fact, a monoclonal antibody raised against P. malariae CSP does recognize sporozoites of Plasmodium brasilianum, a nonhuman primate plasmodial parasite, but not sporozoites of other human plasmodia (13). Moreover, sera from patients infected with P. falciparum do not recognize other human plasmodia (32).

In most areas where malaria is endemic, two or more human plasmodial species coexist (7). Therefore, indigenous populations are exposed to infection with different plasmodia from birth onwards. Evidence suggests that previous exposure to P. vivax or P. malariae leads to lower-than-expected rates of morbidity and mortality following exposure to P. falciparum (7, 22, 37-39, 58). Maitland and colleagues suggested that these outcomes are related to cross-species immunity (38), so an immune response induced by P. vivax and/or P. malariae may somehow affect the course of subsequent infection with P. falciparum. In addition, studies in the 1930s showed that simultaneous or sequential inoculation of blood containing erythrocytes infected with P. falciparum and P. vivax into human volunteers led to suppression of P. vivax blood-stage parasitemia and an increase in relapsing episodes (4-6). Moreover, sera from patients exposed to P. vivax do recognize several proteins in P. falciparum asexual stages, but not in gametes or zygotes (30). Observations in Papua New Guinea, where different plasmodia coexist, suggested that immunity to P. vivax develops much more quickly than immunity to P. falciparum (41), an indication that each species elicits primarily non-cross-reactive immune responses. Other field-derived data from areas where two or more human plasmodia coexist are generally inconsistent and do not allow firm conclusions to be drawn concerning naturally acquired immunity and its effect on cross-species protection.

Cross-protection against one plasmodial species induced by asexual blood stages of another has been observed in both avian (48, 60) and murine (16, 26, 33, 40) models. Primary infection with P. chabaudi chabaudi AS-infected erythrocytes led to protection against mortality and lower parasitemia levels after a challenge infection with a lethal strain of P. yoelii (33). Mice infected with blood-stage P. berghei parasites and later reinfected with P. yoelii-, P. vinckei-, or P. chabaudi-parasitized erythrocytes were partially protected against mortality (26, 40). Coinfection with P. berghei ANKA and P. yoelii (but not P. vinckei) led to inhibition of development of cerebral malaria and to abolition of CD8⁺ T cells in brain blood vessels (65). As seen with exoerythrocytic stages, cross-species protection with plasmodial parasites that infect different hosts has also been observed (10, 69). Sera from mice highly infected with P. chabaudi (parasitemia, ≥50%), for example, inhibit the growth of the asexual blood stages of P. falciparum in vitro (10). Recently, a family of proteins with domains conserved between P. falciparum and P. yoelii, called merozoite released soluble particles, was described (69). These antigens induce a high level of cross-species reactivity, and mice immunized with P. falciparum merozoite released soluble particles exhibit a ∼50-fold decrease in peak parasitemia when they are challenged with P. yoelii.

CONCLUSIONS

Past observations based on experimental (16, 26, 33, 40, 48) and clinical studies (4-6, 11, 15, 27) led us to conclude that there is at least a degree of cross-species immunity within a host coinfected by two or more plasmodial species. Field data, however, are inconclusive on this issue (22, 37-39, 41). The fact that such immunity can indeed be induced experimentally via immunization with live attenuated parasites has been demonstrated with P. berghei RAS and p36p⁻ GAP (19, 45b).

Developing a vaccine for malaria is one of the most sought-after goals in biomedical research. So far, protein subunit- and/or DNA-based vaccines have had only limited success (24, 42), while immunization strategies using P. falciparum RAS have, until very recently at least, been considered a practical impossibility due to technical, clinical, and logistical hurdles (36). The reproducibility and safety of the irradiation process are considered two of the most important and contentious issues, since an overdose kills the sporozoites, with a resulting failure to generate protective immunity, while underirradiated parasites may remain infective (59, 62). The possibility that these hurdles may yet be overcome is evident from the large quantities of aseptic, metabolically active, nonreplicating P. falciparum RAS already produced (36; http://www.sanaria.com). This RAS-based vaccine, produced under good manufacturing practice, is due to enter clinical trials in 2008 (S. L. Hoffman, personal communication). The possibility of replacing RAS with GAP potentially represents a significant step in enhancing the safety of such a live whole-organism vaccine. Orthologues of the different GAP (p36⁻, p36p⁻, uis3⁻, uis4⁻) in P. falciparum and P. vivax have already been described (28, 39a, 61), suggesting that it should be possible to perform this type of genetic modification in human plasmodia, thus avoiding the known hazards of radiation. Another way to circumvent this safety issue might be to use this RAS-based technology not with P. falciparum but with a less virulent species, such as P. vivax or even P. malariae. One of the safety issues would thus be overcome (since any disease-related aspects would be of much less consequence than those with P. falciparum); clearly, if a vaccine based on any one of these species can be shown to be effective against any or all of the others, there would be enormous economic advantages. Manufacturing costs would be lower, since for the price of production of a vaccine for one species, vaccines for the other species could be developed. If a vaccine based on cross-species immunity does not prevent P. falciparum infection per se but nevertheless proves to be effective in preventing mortality, morbidity, and disease-related symptoms, it would still represent a significant achievement that should be strongly encouraged. The argument that the human plasmodial species are phylogenetically distantly related (20) and that therefore no degree of cross-species protection can be expected, can be countered by evidence from rodent models using live attenuated sporozoites of plasmodia that are also considered to be distantly related (46) but that are nevertheless still able to induce cross-species protection (19, 45b). Several advances in in vitro culture methods have been reported recently (30, 55). These advances may allow evaluation of such vaccines in vitro at an earlier stage than is currently possible, with potential savings in both time and resources for development decisions. It has been suggested that a single-species vaccine for P. falciparum could encourage infection and disease due to other malarial parasites that are usually suppressed by P. falciparum (7). In the past, during malaria control programs for example, reduction of P. falciparum infections was found to be followed by an increase in P. malariae cases (8). The development of a vaccine that is effective against multiple species would obviously be expected to preclude such issues, since it should act against all plasmodia infecting a given individual.

In conclusion, then, we consider it strategically important that the possibility of induction of cross-species immunity should be considered and evaluated using the whole-organism approach to malaria vaccine design. With this in mind, further research on the immunological mechanisms involved in induction and maintenance of heterologous antiplasmodial protection by attenuated parasites is clearly warranted.