|Home | About | Journals | Submit | Contact Us | Français|
Malaria vaccines based on ookinete surface proteins (OSPs) of the malaria parasites block oocyst development in feeding mosquitoes and hence disrupt the parasite life cycle and prevent the disease from being transmitted to other individuals. To investigate whether a noninvasive mucosal vaccination regimen effectively blocks parasite transmission in vivo, Plasmodium yoelii Pys25, a homolog of the Pfs25 and Pvs25 OSPs of Plasmodium falciparum and Plasmodium vivax, respectively, was intranasally (i.n.) administered using a complement-deficient DBA/2 mouse malaria infection model, in which a highly elevated level of oocysts develops in feeding mosquitoes. Vaccinated mice developed a robust antibody response when the vaccine antigen was given together with cholera toxin adjuvant. The induced immune serum was passively transferred to DBA/2 mice 3 days after infection with P. yoelii 17XL, and Anopheles stephensi mosquitoes were allowed to feed on the infected mice before or after serum transfusion. This passive immunization completely blocked oocyst development; however, immune serum induced by the antigen or adjuvant alone did not have such a profound antiparasite effect. Further, when i.n. vaccinated mice were infected with the parasite and then mosquitoes were allowed to directly feed on the infected mice, complete blockage of transmission was again observed. To our knowledge, this is the first time that mucosal vaccination has been demonstrated to be efficacious for directly preventing parasite transmission from vaccinated animals to mosquitoes, and the results may provide important insight into rational design of nonparenteral vaccines for use against human malaria.
Malaria is one of the most important infectious diseases, and the levels of mortality and morbidity are high, especially among children in developing countries in Africa, Asia, and South America. Implementation of malaria control measures, such as antimalaria drug chemotherapy and insecticide-treated bed nets, has made a significant contribution to reducing the incidence of malaria in many parts of the world. However, these control measures may not be sufficient, and therefore new tools, including vaccines, should be included in a new malaria control campaign for local elimination and final eradiation of malaria from the globe (7). A promising strategy to counteract global malaria endemicity is to develop highly efficacious vaccines, and several promising candidates have been intensively investigated (7, 20); vaccines targeting asexual stages (i.e., sporozoite, hepatic, and erythrocytic stages) are designed to prevent infection and reduce disease severity, while vaccines that target the sexual stage, in which the parasite undergoes sporogonic development in anopheline mosquitoes, prevent vector-mediated transmission of the parasite from person to person (4, 8, 14, 17, 25). Although transmission-blocking vaccines do not directly prevent infection, they reduce parasite infectivity for the vector and consequently lower the mosquito infection rate and the frequency of transmission to humans. In addition, this strategy is believed to be particularly useful for controlling escape of mutants from vaccines designed based on antigens expressed at an asexual stage; therefore, transmission-blocking vaccines are increasingly being considered indispensable components of malaria vaccine strategies and are key components of malaria elimination (10, 11).
Studies on rodent and human malaria concluded that an effector mechanism that is pivotal for blocking transmission is induction of antigen-specific serum antibodies in a vaccinated host, from which female mosquitoes, when they bite to obtain blood meal, coingest gametocyte pairs together with the induced antibodies (2, 3, 5, 9, 18, 24). The ingested antibodies seem to be stable in the mosquito midgut, at least in the time frame within which the transmitted gametocytes develop into ookinetes.
Parasite antigens expressed later at the postfertilization stage in the mosquito midgut, such as ookinete surface proteins (OSPs), including Pfs25 and Pvs25 from Plasmodium falciparum and Plasmodium vivax, respectively, are particularly important vaccine targets because they are likely to be concealed immunologically, if not concealed completely, from the mammalian host's immunosurveillance system, which suggests that there is a reduced driving force to produce the antigenic variations often observed for antigens expressed at prefertilization stages (6, 12, 23, 28). In addition, several recent studies indicated that in the malaria life cycle the ookinete-to-oocyst transition stage is one of the most vulnerable stages of parasite development, making the postfertilization stage of sporogonic development an ideal target for antitransmission vaccines.
The vast majority of pathogens invade through mucosal tissues and therefore can be controlled effectively by mucosal vaccines rather than parenteral vaccines. Notwithstanding the great merit of mucosal vaccines, most vaccines in use today are delivered parenterally (subcutaneously [s.c.] or intramuscularly). In spite of many arguments against the concept that vaccines against arthropod vector-borne human pathogens, such as malaria parasites, could be designed based on mucosal delivery, recent studies performed by us and other workers demonstrated that mucosal vaccines could be efficacious for prevention of arthropod-transmitted infections, because mucosal administration of foreign antigens mixed with a potent mucosal adjuvant, such as cholera toxin (CT), can induce strong systemic immunity (2, 3, 13). Here we extended our previous studies to test our hypothesis that the malaria OSPs are sufficiently immunogenic when they are administered by the intranasal (i.n.) route in the presence of a mucosal adjuvant, which should in theory effectively block parasite transmission to feeding mosquitoes when both passive and active vaccination regimens are used.
Seven-week-old female DBA/2NCrj (DBA/2) mice were purchased from Japan SLC (Tokyo, Japan). Complement C5-deficient DBA/2 mice were used for live mosquito-feeding experiments, because the highly elevated levels of oocysts that developed in the mosquito midgut were useful for evaluation of transmission-blocking vaccine efficacy (26).
Mice were i.n. vaccinated once a week for 4 weeks with 25 μg of yeast-derived recombinant Plasmodium yoelii Pys25 synthesized and purified like Pvs25 as described previously (14) in the absence or presence of 1 μg of CT (Sigma-Aldrich). As a control, a group of mice were vaccinated with 1 μg of CT alone. For passive vaccination experiments, DBA/2 mice were intravenously vaccinated with 0.5 ml of pooled immune sera derived from mice vaccinated i.n. with Pys25 plus CT, with Pys25 alone, or with CT alone.
For ELISA of vaccine-induced immune sera, a flat-bottom 96-well microtiter plate (Immulon 4; Dynex Technology Inc., Chantilly, VA) was coated with recombinant Pys25 (0.5 μg/well in bicarbonate buffer, pH 9.6) and blocked with 1% skim milk in Tris-buffered saline containing 0.05% Tween 20. Immune sera serially diluted with the blocking buffer were applied to wells in duplicate (100 μl/well) and incubated for 2 h at 37°C, which was followed by addition of alkaline phosphatase-conjugated anti-mouse antibody for immunoglobulin (Ig) isotype and IgG subclass analysis. The alkaline phosphatase substrate (p-nitrophenyl phosphate [Sigma-Aldrich]) was added, and the absorbance at 490 nm was determined with a microplate reader (Bio-Rad Laboratories). The antibody concentration was determined based on known amounts of mouse Igs used as a standard. The statistical significance of differences in antibody concentration or absorbance was determined by Student's t test.
For analysis of the parasite-killing effect of i.n. vaccination-induced immune sera, mice were intraperitoneally inoculated with 106 peripheral red blood cells that had been infected with P. yoelii strain 17XL, and the infected mice were maintained for 3 days until the level of parasitemia reached 9 to 10%, which was determined by microscopic examination of Giemsa-stained thin blood smear preparations. Then approximately 100 Anopheles stephensi mosquitoes that had been starved overnight were allowed to obtain a blood meal from the infected mice either before or 1 h after intravenous injection of immune sera that had been prepared from mice 1 week after the last i.n. vaccination with Pys25 plus CT, with Pys25 alone, or with CT alone. Fully engorged mosquitoes were maintained at 24°C for 1 week by giving them water containing 1.5% fructose and 1.5% sucrose. For each experimental group, mosquitoes were dissected, and their midguts were examined with a light microscope to count the number of oocysts.
For analysis of the direct parasite transmission-blocking efficacy of i.n. vaccination, mice vaccinated with Pys25 plus CT, with Pys25 alone, or with CT alone were infected as described above with the parasite 1 week after the last vaccination, and then mosquitoes were allowed to feed directly on the infected animals; this was followed by enumeration of the oocysts that developed.
The statistical significance of differences in the numbers of oocysts was determined by the Kruskal-Wallis test or the Wilcoxon-Mann-Whitney U test by using the software JMP (SAS Institute Inc.).
A significant level of specific serum IgG and IgM antibodies (mainly IgG) was induced in DBA/2 mice by i.n. vaccination with Pys25 plus CT (14,147 ± 4,241 μg/ml) but not by i.n. vaccination with Pys25 alone or CT alone (Fig. (Fig.1a,1a, upper panel). Oral inoculation of 50 μg of Pys25, however, did not induce an antibody response even in the presence of 10 μg CT (data not shown). IgG1 was found to be the predominant serum IgG subclass, and almost no IgG2a was detected in mice vaccinated with Pys25 plus CT, an indication of the Th2 type of immune response induction (Fig. (Fig.1a,1a, lower panel). Low but detectable levels of Pys25-specific serum IgA and IgE were seen in the group vaccinated with Pys25 plus CT but not in the group vaccinated with Pys25 alone or CT alone (Fig. (Fig.1b).1b). Similar humoral immune responses were observed when outbred ddy mice were used for the immunization experiments (data not shown).
To evaluate the parasite transmission-blocking effect of the induced immune sera in vivo, A. stephensi mosquitoes were allowed to obtain a blood meal from DBA/2 mice that had been infected with P. yoelii 17XL before or after passive transfer of the immune sera of mice vaccinated i.n. with Pys25 plus CT, with Pys25 alone, or with CT alone as described in Materials and Methods. For all three immunization groups, large numbers of oocysts were observed in the mosquito midgut when the mosquitoes were allowed to feed before the immune sera were transferred (median for CT alone, 346 oocysts; median for Pys25 alone, 302 oocysts; median for Pys25 plus CT, 311 oocysts) (Fig. (Fig.2a).2a). In contrast, when the mosquitoes received the blood meal after the immune sera were transferred, oocyst formation was completely blocked in the group vaccinated with Pys25 plus CT, but not in the group vaccinated with CT alone (median, 109 oocysts) or with Pys25 alone (median, 65 oocysts). Although we do not know why the CT or Pys25 immune serum had a significant parasite-killing effect (for CT, 346 oocysts versus 109 oocysts; for Pys25, 302 oocysts versus 65 oocysts), no mosquitoes completely lacked oocysts when they were given CT or Pys25 immune serum (Table (Table1).1). The results demonstrated that i.n. vaccination with Pys25 plus CT induced antibodies which confer complete transmission-blocking immunity when a passive vaccination regimen is used.
Next, to evaluate the direct mucosal vaccine efficacy of Pys25, mosquitoes were allowed to obtain a blood meal directly from parasite-infected mice that had been vaccinated as described in Materials and Methods. The results demonstrated that vaccination with Pys25 plus CT completely blocked oocyst development, as we observed in the passive vaccination experiment, while significant numbers of oocysts were observed in mosquitoes that fed on CT-vaccinated mice (median, 269 oocysts) or Pys25-vaccinated mice (median, 170.5 oocysts) (Fig. (Fig.2b).2b). Vaccination with Pys25 alone had a weak but significant transmission-blocking effect compared with the effect observed for CT-vaccinated mice, suggesting that i.n. vaccination with the recombinant antigen alone might have some efficacy, even though the antibody levels for these two groups were not significantly different (Fig. (Fig.1).1). The oocyst prevalence was 100% for all vaccination regimens that we tested except both the passive and active Pys25-plus-CT regimens, for which the oocyst prevalence was 0% (Table (Table1).1). On the basis of our results, we concluded that i.n. vaccination with the malaria OSP was very efficacious when this OSP was combined with a mucosal adjuvant to block parasite transmission to mosquitoes.
Although most vaccines in use today are administered s.c. or intramuscularly, the advantages of mucosal vaccines are indisputable; they result in local immunity as well as systemic immunity, which, in general, is hard for parenteral vaccines to induce, and they provide a first line of defense against many infections that occur at or emanate from mucosal surfaces. They could prevent transmission of blood-borne pathogens by reuse of syringes; they may be safer and more cost-effective and thus have advantages for developing countries; and they are painless and therefore likely to be readily tolerated by small children and individuals with needle phobia (16). Although not all of the advantages attributed to mucosal vaccines mentioned above are directly relevant to the design of vaccines against malaria, and although there are some intrinsic technical difficulties that cannot be circumvented by development of effective mucosal vaccines (7), evaluation of the concept of designing mucosal vaccines for nonmucosal pathogens seems to be worthwhile.
To investigate malaria parasite OSP-based mucosal vaccines, we previously demonstrated that for two types of human malaria (P. falciparum malaria and P. vivax malaria) experimentally induced mouse immune sera specific for Pfs25 and Pvs25 (homologues of rodent Pys25) were very effective in blocking parasite transmission from patients' parasitized blood to mosquitoes in a membrane feeding assay (2, 3). In the present study we demonstrated that in a rodent malaria infection model, OSPs were immunogenic when they were administered i.n. (with levels of antigen-specific Igs reaching 15 mg/ml), and the induced immune serum was very effective in blocking parasite transmission. Most importantly, however, we demonstrated that vaccination directly prevented the transmission of a parasite from vaccinated animals to feeding mosquitoes. To our knowledge, this is the first demonstration that mucosal vaccination with malaria OSPs can directly prevent malaria transmission to mosquitoes in vivo. In the DBA/2 strain of mice lacking a component of the complement system, the number of oocysts formed is significantly increased in feeding mosquitoes (26); however, anopheline mosquitoes collected in field are usually not as heavily infected, and a single oocyst is commonly detected. Therefore, a more moderate antibody level may confer effective transmission-blocking immunity in humans (21).
Malaria vaccines targeting hepatic and erythrocytic stages suffer from antigenic variations mainly due to selection pressure from the host immune system. However, antigens expressed at a parasite sexual stage, such as OSPs, are immunologically concealed from the host immune system, and hence the chance that antigenic variations occur may be low. Indeed, OSPs of P. falciparum and P. vivax were shown to have minimal antigenic variations even in field isolates collected from remote regions of the world (19, 27). This is an important characteristic of ideal vaccines. On the one hand, sexual-stage antigens have disadvantages such as (i) the absence of an infection-induced booster effect and the resulting long-term immunity and (ii) the absence of direct protection of vaccinees from infection. Therefore, it is believed that a vaccine candidate should have multiple components and that at least one component should be a sexual-stage antigen (7). In such a vaccine formulation, preerythrocytic and/or erythrocytic antigens may function cooperatively with sexual-stage antigens for prevention of or reduction of infection and parasite transmission.
Mucosal administration, such as i.n. or oral administration, unlike parenteral immunization, of nonreplicating inert antigens with CT tends to induce Th2-type immunity, which is characterized by predominant induction of serum IgG1, induction of local secretory IgA, and in some cases induction of serum IgE in mouse models. Unlike what happens in other infectious diseases, which require induction of cell-mediated immunity (22), serum antibody, regardless of the Ig isotype, seems to be the predominant, if not only, protective arm of immunity that blocks malaria transmission. We do not know the mechanism of action of Pys25-specific antibodies in blocking parasite development in the midgut of a feeding mosquito, but binding of antibodies to the zygote surface and subsequent prevention of parasite development into the ookinete may be the most important blocking mechanism (14, 25). This antibody binding may occur within the midgut, and this may be independent of Ig isotypes. Thus, although IgE antibody is not the major antibody isotype present in a vaccinated host, it may contribute to blocking transmission. However, the induction of serum IgE may potentially lead to an allergic response in vaccinated individuals, and the data shown in Fig. Fig.1b1b are relevant to this issue. Another important issue that needs to be considered is the duration of protective antibodies. Recent findings relevant to the present work indicated that when P. vivax transmission-blocking vaccine candidate Pvs25 was injected s.c. with incomplete Freund's adjuvant into BALB/c mice, it induced a strong antigen-specific serum IgG response that was maintained for more than 6 months (our unpublished data). i.n. vaccination with Pvs25 plus CT induced a level of serum IgG comparable to that induced by s.c. vaccination formulated with incomplete Freund's adjuvant, but the level gradually decreased over 6 months. However, we found that i.n. vaccination with Pvs25 plus CT was generally more potent based on the magnitude and duration of the specific serum IgG response than s.c. vaccination with Pvs25 formulated with aluminum hydroxide (unpublished data).
In this study we used CT as a mucosal adjuvant; however, the use of CT for humans is hampered by the toxicity of this compound. Also, as mentioned above, issues related to the potential allergic response and the duration of antibodies need special consideration. Fortunately, however, nontoxic and thus safer adjuvants, but adjuvants that are as effective as CT, are being developed, making a mucosal malaria vaccine a feasible goal (1, 15). For example, we recently found that when a nontoxic subunit of CT, CTB, was fused to malaria OSP, it was efficacious by both the mucosal and s.c. routes for blocking parasite transmission (unpublished data). Thus, if the mucosal transmission-blocking vaccine efficacy data obtained with this rodent infection model can be reproduced in human clinical trials with guaranteed safety, OSP antigens formulated as noninvasive vaccines may become a powerful tool for use against human malaria.
This work was supported by the Program of Founding Research Centers for Emerging and Reemerging Infectious Diseases of the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan. It was also supported in part by Grant-in-Aid for Scientific Research 18390129 from MEXT, Japan.
Editor: W. A. Petri, Jr.
Published ahead of print on 14 September 2009.