Protection against malaria can be achieved by induction of a strong CD8+ T-cell response against the Plasmodium circumsporozoite protein (CSP), but most subunit vaccines suffer from insufficient memory responses. In the present study, we analyzed the impact of postimmunization sporozoite challenge on the development of long-lasting immunity. BALB/c mice were immunized by a heterologous prime/boost regimen against Plasmodium berghei CSP that induces a strong CD8+ T-cell response and sterile protection, which is short-lived. Here, we show that protective immunity is prolonged by a sporozoite challenge after immunization. Repeated challenges induced sporozoite-specific antibodies that showed protective capacity. The numbers of CSP-specific CD8+ T cells were not substantially enhanced by sporozoite infections; however, CSP-specific memory CD8+ T cells of challenged mice displayed a higher cytotoxic activity than memory T cells of immunized-only mice. CD4+ T cells contributed to protection as well; but CD8+ memory T cells were found to be the central mediator of sterile protection. Based on these data, we suggest that prolonged protective immunity observed after immunization and infection is composed of different antiparasitic mechanisms including CD8+ effector-memory T cells with increased cytotoxic activity as well as CD4+ memory T cells and neutralizing antibodies.
There is now goodprogress both in the prevention and the treatment of malaria that leads to decreasing numbers of malaria cases and fatalities; nevertheless malaria is still one of the biggest challenges for global public health management. There are around 250 million cases per year worldwide with the vast majority of these in the African region and it is estimated that malaria accounts for 900 000 deaths annually . Therefore a vaccine is still the ultimate goal of malaria research.
The development of immunity and immunological memory to Plasmodium infections in malaria endemic areas has been intensively studied . Children in the age of 6 months to 5 years are at the highest risk of being infected with Plasmodium and developing severe disease. In the first years of life, an anti-disease immunity is acquired that leads to high-density infections with mild symptoms. In contrast, the acquisition of anti-parasite immunity takes years to decades, although transmigrant studies show that adults may acquire immunity more rapidly than children [3, 4]. It is supposed that anti-parasite immunity is based mainly on antibodies recognising variant surface antigens on the membranes of infected erythrocytes whereas immune responses against preerythrocytic forms seem to play a minor role [5, 6]. Naturally acquired immunity is never sterile but enables adults in endemic areas to be infected without clinical symptoms and with low parasite densities. It is still under debate whether malaria parasites have evolved specific immune evasion mechanisms that interfere with memory formation. There are several reports of previously immune immigrants losing their immunity to malaria after a certain time of residence in a nonendemic area; however, systematic surveys imply that previously immune adults residing in nonendemic areas maintain a lower risk of severe and fatal disease [7-9].
It was shown by using irradiated or genetically attenuated sporozoites that sterile protection can be achieved by the induction of a T-cell response directed against infected hepatocytes [10-13]. In addition, a liver-stage-specific vaccine has the advantage that it could prevent the pathology associated with the blood-stage of the disease. Many vaccine trials are based on subunit vaccines that induce CD8+ T cells against liver-stage antigens. The best characterized and most dominant antigen is the circumsporozoite protein (CSP) that was shown to induce potent T-cell as well as humoral responses upon vaccination with irradiated sporozoites [11, 14]. Sterile protection against sporozoite challenge can be transferred to naïve mice by CSP-specific CD8+ T cells [15-18]. There are several experimental vaccines described that successfully induce CSP-specific CD8+ T cells [19-22].
In the present study we employed our previously described heterologous prime/boost vaccination strategy that is based on the induction of CD8+ T cells recognising a CSP-derived epitope of Plasmodium berghei presented by MHC class I . Primary immunization is achieved with recombinant Salmonella. The strain Salmonella typhimurium SB824/pST-TB was engineered to express a fusion molecule of the Yersinia enterocolitica protein YopE and the CD8+ T-cell epitope CSP245–253 of P. berghei. YopE is mediating thereby the delivery of the CSP fragment into the MHC class I presentation pathway via the type III secretion system. For boosting, we made use of a recombinant Bordetella pertussis adenylate cyclase toxoid (ACT) as antigen carrier. ACT is a cell-invasive toxoid that was shown to target professional antigen-presenting cells due to its high affinity binding to CD11b/CD18 on myeloid dendritic cells (DCs) . ACT was shown to carry various inserted passenger proteins within its N-terminal AC domain without losing the capacity to translocate this domain directly across the cytoplasmic membrane into the cytosol of target cells [25-27]. The recombinant vaccine ACT-CSP is genetically detoxified and contains within its catalytic domain a 24-mer of P. berghei CSP including the CD8+ T-cell epitope CSP245–253 together with its flanking amino acids . This vaccination regimen induces a high number of CSP-specific CD8+ T cells producing IFN-γ and TNF-α that show a strong cytotoxicity in vivo. Subsequently to this heterologous prime/boost vaccination, mice are nearly completely protected when challenged with P. berghei ANKA sporozoites.
In the present study, we analyzed the longevity of the vaccine-induced protective immunity and dissected how a sporozoite challenge of vaccinated mice alters the quantity and quality of liver stage-specific immunity. We show that upon sporozoite challenge protective humoral immunity was induced, but which was not sufficient to mediate sterile protection. The number of CSP-specific CD8+ memory T cells was not changed significantly after sporozoite challenge; however, these T cells displayed improved cytotoxic function and were essential for sterile protection. CD4+ memory T cells were seemingly primed upon challenge as well but they played a minor role for sterile protection. We conclude that a sporozoite challenge of vaccinated mice induces multifaceted immune responses including sporozoite-specific antibodies and CD4+ T cells; however, CD8+ T cells still play the central role for sterile protection.
Immunization with S. typhimurium SB824/pST-TB followed by ACT-CSP boost induces stable memory
In our previous study, we described a heterologous prime/boost immunization strategy that induces a strong CSP-specific CD8+ T-cell response . Primary immunization is done with three doses of the recombinant S. typhimurium AroA mutant strain SB824/pST-TB expressing the CD8+ T-cell epitope SYIPSAEKI from P. berghei CSP. As boost immunization the recombinant B. pertussis adenylate cyclase toxoid ACT-CSP is used that also contains the P. berghei CSP CD8+ T-cell epitope (Fig. 1A). This immunization strategy induces high numbers of CSP-specific CD8+ T cells as shown by MHC-I pentamer staining (Fig. 1A and Supporting Information Fig. 1) that produce IFN-γ and TNF-α upon antigenic restimulation in vitro (Fig. 1B and Supporting Information Fig. 2). In the present study, we further analyzed the duration of the CSP-specific T-cell response. We found that heterologous prime/boost immunization induced a CSP-specific CD8+ memory T-cell population that remained stable after 8 weeks postimmunization (Fig. 1A). These memory T cells were capable of producing IFN-γ and TNF-α upon antigenic restimulation (Fig. 1B).
Sporozoite challenge after immunization leads to prolonged durability of protection
Our heterologous prime/boost vaccination regimen consisting of S. typhimurium SB824/pST-TB and ACT-CSP was already shown to induce nearly complete protection with only very few vaccinated mice developing blood-stage disease . To test the longevity of the protective immune response mice were challenged at different time points after boost immunization (Table 1). Protection rates declined from 96% when mice were challenged on day 7 after vaccination to 16% when challenged three months later (Table 1). Since people living in malaria-endemic areas are frequently infected with Plasmodia, we wanted to investigate whether a sporozoite challenge of vaccinated mice could further enhance and prolong protective immunity. Therefore, we vaccinated mice using a regimen as depicted in Figure 1A followed by a sporozoite challenge at day 7 postboost immunization and made a second challenge 12 weeks after the first one. These mice showed a significantly elevated protection rate of 72% compared with 16% protection of mice that were only immunized (Table 1). Thus a sporozoite challenge of vaccinated mice significantly prolonged protective immunity.
Table 1. A heterologous prime/boost vaccination with S. typhimurium SB824/pST-TB followed by ACT-CSP induces only short-lived sterile protection against P. berghei, which can be prolonged by sporozoite challenge
BALB/c mice were immunized with three oral doses of S. typhimurium SB824/pST-TB followed by i.p. ACT-CSP boost 5 weeks later.
At the indicated time points after boost immunization mice were challenged i.v. with 1000 P. berghei ANKA sporozoites. Animals were scored as protected when no infected erythrocytes were detectable in blood smears until day 14 i.p. and all naïve control mice developed a parasitaemia. Variations in the course of parasitaemia were not considered.
Some of the mice that were challenged at day 7 after boost immunization received a second challenge 12 weeks later.
Sporozoite challenge after immunization does not substantially boost the CSP-specific T-cell population
Next, we were interested whether the increased protection of vaccinated mice after sporozoite challenge is due to an enhanced CSP-specific CD8+ T-cell memory response. Therefore, we vaccinated and challenged mice as depicted in Figure 2A. Three months after boost immunization we determined the number of CSP-specific CD8+ T cells within blood, spleens, and livers of the mice. In comparison with mice that were only immunized, the numbers of CSP-specific T cells were only slightly enhanced in the blood of animals that received a challenge but there was no difference in spleen and liver (Fig. 2A). In addition, the numbers of CSP-specific IFN-γ-, TNF-α-, and IL-2-producing T cells were also only slightly enhanced (Fig. 2B). Therefore, we conclude that improved protection after the first challenge cannot be simply explained by enhanced numbers of CSP-specific CD8+ T cells.
Correlation of CSP-specific CD8+ T cells with protection
Mice that are immunized by a heterologous prime/boost immunization with three doses of S. typhimurium SB824/pST-TB and ACT-CSP are almost completely protected whereas the same vaccination regimen with only one dose of Salmonella was shown to induce only partial protection of 70% . Therefore, we looked for a correlation of the number of CSP-specific T cells with protection when mice were immunized by the two different vaccination regimens (Fig. 3A). Seven days after the last immunization all mice were challenged with P. berghei sporozoites. One day before challenge the numbers and cytokine production of CSP-specific T cells in the blood were analyzed. We found a significant correlation between either the numbers of CSP-specific CD8+ effector T cells or the numbers of CD8+ IFN-γ+TNF-α+ T cells and protection (Fig. 3A). We next wanted to know whether the numbers of CSP-specific CD8+ memory T cells also correlate with long-term protection. Therefore, we immunized mice and challenged one group of animals 7 days after boost immunization whereas one group of vaccinated mice was left untreated as described in Figure 3B. Three months after boost immunization, we challenged all mice and analyzed one day before challenge CSP-specific CD8+ memory T cells in the blood. In contrast to the clear correlation seen 7 days after boost immunization, no correlation between protection and the numbers or cytokine production of CSP-specific CD8+ memory T cells was found (Fig. 3B). This was independent of whether the mice had been challenged after vaccination or not. These findings suggest that other immune mechanisms beyond the mere number of CSP-specific CD8+ T cells contribute to the improved long-term protective immunity after challenge of vaccinated mice.
Induction of sporozoite-specific antibodies upon sporozoite challenge
It is known that immunization with attenuated sporozoites induces humoral immune responses against Plasmodium antigens  and it was shown that repeated sporozoite challenges of vaccinated mice induce sporozoite-specific antibodies as well . Therefore, we evaluated the contribution of sporozoite-specific antibodies to prolonged protection observed after sporozoite challenge. To this end, mice were immunized by our heterologous prime/boost vaccination regimen and challenged 7 days after boost and additionally 3 months after boost. Twenty-one days after first and second challenge serum samples were taken. To detect total sporozoite-specific antibodies P. berghei sporozoites were isolated from Anopheles stephensi salivary glands and air-dried onto glass slides. After fixation sporozoites were stained with mouse sera in serial dilutions and anti-mouse pan-IgG secondary antibody (Fig. 4A and B). Total sporozoite-specific titer was defined as maximum dilution that still gave a positive staining in comparison to naïve control serum. All sera obtained from mice after first challenge stained sporozoites with a mean titer of 1800. After second challenge titers increased substantially and the mean titer was around 8000 (Fig. 4B). Of note, the vaccine itself contains only the minimal sequence of the described CD8+ T-cell epitope of CSP and the vaccination alone does not induce any sporozoite-specific antibody response. Additionally, we performed a CSP-specific ELISA from mouse sera obtained after first and second challenge. Therefore, a recombinant P. berghei CSP was used that contains a shortened repeat region to allow efficient expression in Escherichia coli. As already seen by immunofluorescence CSP-specific IgG titers increased significantly after each challenge with mean titers of 160 and 700 after first and second challenge, respectively (Fig. 4C). The different titers determined by immunofluorescence to whole sporozoites and by CSP-specific ELISA might reflect the presence of antibodies with different specificity within sera. On the other hand side, it is conceivable that not all CSP-specific antibodies could be detected due to the deletion within the repeat region of the recombinant CSP what might result in the absence of certain linear or conformational B-cell epitopes.
Protective capacity of mouse serum after challenge
In order to evaluate the protective potential of sporozoite-specific antibodies within the serum of mice induced by sporozoite challenge, we first performed in vitro sporozoite inhibition assays. Therefore, HepG2 cells were infected with P. berghei sporozoites that were preincubated with serum obtained from mice after first or second challenge or naïve control serum in a 1:10 dilution. Relative inhibition of sporozoite invasion was calculated in relation to the number of infected cells without serum preincubation. Serum obtained from mice 21 days after first challenge significantly inhibited parasite invasion by 40% in comparison with naïve serum (Fig. 5A). This inhibitory capacity was further increased to over 60% inhibition with serum obtained 21 days after second challenge. Next, the protective capacity of serum was determined in vivo. To this end, serum was obtained from mice 21 days after either first or second challenge and transferred into naïve mice. Mice were challenged subsequently with P. berghei sporozoites. Monitoring of blood stage parasitemia did not reveal any protective effect of a serum transfer (Fig. 5B) despite substantial amounts of sporozoite-specific IgG.
Contribution of CD8+ and CD4+ T cells to prolonged protection after first challenge
Next, we were interested whether the sporozoite-specific antibody response induced by the first challenge or potentially induced CD4+ memory T cells would be sufficient to protect mice in the absence of CSP-specific memory CD8+ T cells. To this end, we immunized mice and challenged 7 days and 3 months after boost immunization. One day before the second challenge, a group of mice was depleted of CD8+ T cells by injection of the monoclonal anti-CD8 antibody YTS169. Depletion of CD8+ T cells abrogated sterile protection as shown in Figure 6A; nevertheless anti-CD8-treated mice still showed a significantly reduced parasitaemia in comparison with naïve control mice. This partial protection might either be due to sporozoite-specific antibodies or to CD4+ memory T cells that could have been primed upon first challenge. Therefore, we repeated the above-described experiment but depleted CD4+ T cells by injection of the monoclonal antibody GK1.5 at days –3 and –1 before second challenge. The depletion of CD4+ T cells led to a partial but not complete loss of sterile protection as shown in Figure 6B. These results made clear that antibodies and CD4+ memory T cells contribute to the observed prolonged protection after challenge but both are not sufficient to mediate sterile protection alone.
Functional and phenotypical differences of CD8+ memory T cells after challenge of vaccinated mice
Since CD8+ T cells were absolutely required for sterile protection although the numbers of CSP-specific CD8+ memory T cells were not significantly changed after challenge, we further evaluated functionality of CSP-specific memory T cells. To this end, we performed an in vivo cytotoxicity assay with mice that were immunized according to Figure 1A and challenged afterwards or not. A mixture of control cells (CFSElow/Ag−) and CSP245–253 peptide loaded target cells (CFSEhigh/Ag+) were adoptively transferred. CSP-specific lysis was visualized in vivo by selective reduction of the transferred target cells in relation to control cells after 3 h of circulation within vaccinated or vaccinated and challenged mice in comparison with nonvaccinated mice. Nine weeks after boost immunization, we found a significantly increased rate of CSP-specific lysis in mice that had a sporozoite challenge after immunization compared with that in vaccinated-only mice (Fig. 7A and Supporting Information Fig. 3). Since this might reflect an altered memory subset composition, we additionally analyzed at this time point the expression of different migration and activation markers on CSP-specific memory CD8+ T cells in spleens and livers of mice that were either challenged after vaccination or not (Fig. 7B–D and Supporting Information Fig. 4). The only marker that was found to be differentially expressed when mice had a challenge after vaccination was KLRG1 in the liver (Fig. 7B). There was no difference in CD62L expression arguing for equal proportions of effector and central memory T cells in both groups of mice (Fig. 7C and D, and Supporting Information Fig. 5). The expression level of other markers for the activation status or proliferative potential like CD43 (activation-associated 130 kD glycoform), CD27, CD127, and CD122 was also comparable indicating that the increased cytolytic capacity observed after challenge is not reflected by a substantially altered phenotype of the CSP-specific memory T cells.
In our previous study, we presented a novel vaccination regimen against the liver stage of P. berghei consisting of a recombinant S. typhimurium strain and a recombinant B. pertussis adenylate cyclase both containing the CSP CD8+ T-cell epitope. This heterologous prime/boost immunization was shown to induce a strong effector T-cell response leading to nearly complete protection of vaccinated mice . In the present study, we evaluated the memory response that is induced by this vaccination regimen. We showed that our vaccination regimen induces a stable memory response and that the majority of these CSP-specific memory T cells produce IFN-γ and TNF-α after antigenic stimulation. Nevertheless these memory T cells are not sufficient to completely protect mice against sporozoite challenge 3 months after boost immunization. However, if mice are challenged with sporozoites shortly after vaccination there is a high degree of protection against a secondary challenge 3 months later. This argues for further activation of sporozoite-specific immune responses upon challenge.
We first analyzed CSP-specific CD8+ T cells in the blood of mice 3 months after vaccination or vaccination and challenge and found neither a substantial increase in the numbers of CSP-specific T cells nor a change in cytokine expression. Additionally, there was no correlation between the numbers of CSP-specific memory T cells in the blood and protection against sporozoite challenge. This was in contrast to the situation shortly after vaccination where the number of CSP-specific effector T cells in the blood clearly correlates with protection. These observations argued for a contribution of other immune mechanisms than the mere number of CSP-specific T cells to the prolonged protection observed.
It was described that sporozoite challenges after vaccination induce humoral immune responses . Consistent with this study, we also found sporozoite-specific IgG in the sera of challenged mice and titers increased significantly after each challenge. Sporozoite-specific antibodies in mouse sera showed significant protective capacity in an in vitro assay. However, this protective effect could not be seen in vivo after transfer of serum of mice that were exposed twice to viable sporozoites. Sterile humoral immunity against sporozoites requires very high antibody titers due to the short time span between sporozoite inoculation and hepatocyte infection. With repeated infections the humoral immune response could eventually overcome a certain threshold and completely prevent hepatocyte infection by itself leading to sterile protection. This was demonstrated by Schmidt et al.  who described sterile humoral immunity after five separate sporozoite exposures. The authors concluded that the increasing humoral response would compensate for a waning CD8+ memory T-cell response. However, in our experimental setting the humoral response after two sporozoite challenges was not sufficient, therefore we had a closer look at the CD8+ memory T cells. Based on our data, we conclude that the repeated sporozoite infections rather lead to an improvement of CD8+ T-cell memory.
This is supported by the finding that sterile protection against rechallenge is absolutely dependent on CD8+ T cells despite the presence of protective sporozoite-specific antibodies as shown by depletion of CD8+ T cells. A depletion of CD4+ T cells before rechallenge, however, also led to a partial loss of protection arguing for a priming of CD4+ T cells upon first challenge and a contribution of CD4+ memory T cells to protection later on. However, CD4+ T cells are not sufficient to mediate sterile protection in the absence of CD8+ T cells, whereas sterile protection was still present in few mice after depletion of CD4+ T cells.
Our results indicate that CSP-specific effector-memory T cells after vaccination and challenge are superior to effector-memory T cells induced by vaccination alone. The proportions of effector and central memory T cells as defined by CD62L expression were similar in all mice analyzed but analysis of CSP-specific in vivo cytotoxicity revealed an enhanced cytotoxic activity of memory T cells after vaccination and challenge in comparison with that of memory T cells from vaccinated-only mice. Therefore, we suppose that upon sporozoite challenge a functional maturation of CSP-specific memory T cells takes place that leads to a selective enrichment of high avidity T-cell clones with improved cytotoxic activity. This finding was not associated with dramatic phenotypical changes as the expression of the activation and proliferation markers CD127, CD122, CD27, and CD43 on CSP-specific memory T cells was not altered after challenge. The only marker that was found to be differentially expressed was KLRG1. CSP-specific memory T cells of mice that had been immunized and challenged showed a significantly increased proportion of KLRG1high cells in the liver but not in spleen. KLRG1 is known as a marker of senescence that is expressed by a minor population of CD8+ memory T cells . Its expression was described to be associated with low proliferative but high cytotoxic potential, but whether KLRG1 has a functional relevance for memory T cells or just reflects activation history is still under debate [32, 33]. Our results indicate that repeated antigenic stimulation leads to an alteration of functional characteristics of memory T cells irrespective of their commitment to TEM or TCM lineage and accompanied by only minor phenotypical changes.
In our experimental context antibodies might just complement cellular immunity. In addition there is evidence that antibodies and T cells can act synergistically leading to improved induction and maintenance of antiviral T-cell responses and to enhanced levels of protection against different viral infections [34-36]. With regard to our results one explanation could be that anti-sporozoite antibodies support the maintenance of CSP-specific memory T cells but also facilitate the activation of these memory T cells in case of reinfection with Plasmodium sporozoites. This issue needs to be further investigated.
People in endemic areas develop after repeated infections a nonsterile immunity against malaria  that is supposed to be based on antibodies against blood-stage antigens and IFN-γ made by CD4+ T cells [5, 6]. Antibodies to sporozoites are also present in people living in endemic areas but at very low titers and after years of repeated exposure . However, nonimmune temporary residents in malaria-endemic areas acquire rapidly anti-sporozoite antibodies when blood-stage malaria is suppressed by chemoprophylaxis . It is still a matter of debate how blood stage parasitaemia might influence immune responses against the liver stage. It has been assumed that a blood stage infection might suppress cellular immune responses against liver stage antigens by interference with antigen presentation by DCs [39-41] or by induction of massive apoptosis of Plasmodium-specific T and B cells [42, 43]. In this context, it would be very interesting whether a blood stage infection would interfere with boosting of a preexisting vaccine-induced cellular immune response against CSP or with induction of sporozoite-specific antibodies.
Taken together, the results of our study indicate that natural Plasmodium infections in endemic areas might have a beneficial effect on preexisting vaccine-induced immune responses and that vaccine-induced immune responses might facilitate the acquisition of natural immunity as well.
Materials and methods
Mice and parasites
BALB/c (H-2Kd) mice were bred in the animal facility of the Bernhard Nocht Institute for Tropical Medicine. Animal maintenance and experimental procedures were all approved by the Amt für Gesundheit und Verbraucherschutz of the Hansestadt Hamburg. For all experiments 6–8 weeks old female BALB/c were used. P. berghei ANKA parasites were maintained by alternating passages in Anopheles stephensi mosquitoes and BALB/c mice in the mosquito colony of the Bernhard Nocht Institute for Tropical Medicine. Sporozoites were isolated by manual dissection of salivary glands of infected mosquitoes in PBS between day 18 and 25 after the mosquito had taken an infectious blood meal.
Bacterial strains, growth conditions, and plasmids
Construction of the plasmid pST-TB was outlined in detail previously . Briefly, the plasmid pST-TB encodes SycE, the specific chaperone for Yop E, and the fusion YopE 1–138/CSP 231–260/M45. Translational fusion between YopE 1–138 and CSP 231–260 was constructed by PCR cloning. The cDNA of Plasmodium berghei ANKA strain CSP was used as template DNA for the PCR amplification of CSP 231–260. S. typhimurium strain SB824 was engineered by introducing sptP::kan mutant allele from strain SB237  into the ΔaroA strain SL3261  by P22HTint transduction. pST-TB was transformed into S. typhimurium SB824  by electroporation. S. typhimurium were grown in Luria Bertani medium supplemented with 0.3 M NaCl, pH 7.0, plus ampicillin (100 μg/mL), and kanamycin (50 μg/mL). To prepare Salmonella for immunization of mice, bacteria were grown overnight in Luria Bertani supplemented with 0.3 M NaCl and antibiotics, diluted 1/20 in fresh medium and grown for another 3 h to reach an OD600 of 0.6.
ACT-CSP toxoid construction and purification
The construction of ACT-CSP was described in a previous study . Briefly, the amino acid sequence VRVRKNNDDSYIP SAEKILEFVKQ, which comprises the MHC class I epitope SYIPSAEKI (CSP245–253), was inserted at position 336 into the catalytic domain of the adenylate cyclase of B. pertussis. E. coli strain XL1-Blue (Stratagene) transformed with the appropriate plasmid construct was used for production of the detoxified ACT-CSP. The protein was purified by ion-exchange chromatography on DEAE-Sepharose and hydrophobic chromatography on Phenyl-Sepharose, as described previously . In the final step, the proteins were eluted with 8 M urea, 2 mM EDTA, 50 mM Tris-HCl, pH 8.0, and stored at -20°C. The endotoxin content in the samples was determined by the LAL assay QCL-1000 (Cambrex), according to the manufacturer's instructions and was below 100 EU/mg of purified protein.
Immunization and challenge
Mice were orally immunized with three doses of 5 × 108S. typhimurium SB824/pST-TB at days 0, 3, and 7. Mice were boosted with 20 μg ACT-CSP, which was injected i.p. 5 weeks after Salmonella application. Some mice were immunized by a single dose of S. typhimurium SB824/pST-TB and boosted with 20 μg ACT-CSP 5 weeks later. Seven days, 1 or 3 months after ACT-CSP boost mice were challenged with 1000 Plasmodium berghei ANKA sporozoites i.v. Parasitaemia was determined by staining of blood smears with Wright's stain (Sigma, Taufkirchen, Germany). Some mice were challenged a second time 3 months after ACT-CSP boost with 1000 P. berghei ANKA sporozoites i.v.
For the evaluation of the protective capacity of serum naïve mice received 500 μL of serum i.p. and were challenged 3 h later with 1000 P. berghei sporozoites i.v. The sera were obtained 21 days after either first or second challenge of vaccinated mice.
Isolation of lymphocytes and collection of serum samples
Blood was collected by submandibular bleeding into EDTA-coated or uncoated tubes for isolation of PBL or preparation of serum. For flow cytometry analysis of PBL red blood cells were lysed by addition of ammonium chloride. For preparation of serum blood was allowed to clot for 30 min and centrifuged afterwards for 10 min at 10 000 rpm. Spleens were removed and single cell suspensions were prepared; RBCs were lysed by addition of ammonium chloride. The livers were flushed via the portal vein with 10%FCS/PBS and homogenized through a cell strainer. Suspension was centrifuged 10 min at 380 g and resuspended in pure RPMI 1640. This suspension was mixed with a 30% Nycodenz® solution (Nycomed Pharma AS, Oslo, Norway) and layered under RPMI 1640. After centrifugation for 20 min at 4°C and 1200 g the layer containing liver immune cells was recovered and washed once with 10% FCS/PBS.
Stimulation of cells
PBL were stimulated with CSP245–253 peptide (Proimmune, Oxford, UK) for 5 h at 37°C in the presence of GolgiStop (BD Biosciences, Heidelberg, Germany). After staining of surface markers cells were permeabilized by incubation with Cytofix/Cytoperm (BD Biosciences, Heidelberg, Germany) and intracellular cytokines were stained.
Antibodies, immunofluorescence analysis, and flow cytometry
For in vivo depletion of CD8+ T cells the monoclonal antibody YTS169  (kind gift from H. W. Mittrücker, University Hospital Eppendorf, Hamburg, Germany) was used. Mice were injected once with 500 μg anti-CD8 i.p. 1 day before challenge. Efficiency of depletion was checked by flow cytometric analysis of PBL (data not shown). For in vivo depletion of CD4+ T cells the monoclonal antibody GK1.5 was used. Mice received two i.p. injections of 300 μg anti-CD4 at days –1 and –3 before challenge. Efficiency of depletion was checked by flow cytometry (data not shown).
For determination of IgG titers by immunofluorescence analysis P. berghei sporozoites were fixed with 4% PFA on glass slides. Sporozoites were stained with mouse serum diluted in PBS (1:100–1:40 000) and a secondary anti-mouse pan-IgG FITC conjugate (Jackson ImmunoResearch, 1:250 in PBS) together with DAPI (Sigma, 1 μg/mL).
For intracellular flow cytometry anti-IFN-γ-PE (XMG1.2, eBioscience), anti-TNF-α-FITC (MP6-XT22, eBioscience), anti-IL-2-allophycocyanin (Jes6–5H4, eBioscience), and anti-CD8-PerCP-Cy5.5 (53–6.7, eBioscience) were used. The frequency of CSP-specific T cells was determined using an allophycocyanin -labeled H2-Kd pentamer loaded with the CSP245–253 peptide (Proimmune, Oxford, UK) together with anti-CD8-PerCP-Cy5.5 (53–6.7, eBioscience). For phenotypic analysis anti-CD62L-FITC (MEL-14, BD Pharmingen), anti-CD127-PE (A7R34, eBioscience), anti-CD27-PE (LG.7F9, eBioscience), anti-CD43-FITC (1B11, eBioscience), anti-CD122-Biotin (TM-β1, BioLegend), and anti-KLRG1-Biotin (2F1, BioLegend) together with Streptavidin-PE/Cy7 (BioLegend) were used. Data were acquired on BD™ LSRII or BD FACSCalibur flow cytometer and analyzed by FlowJo software (Tree Star, Ashland, OR).
Cloning, recombinant expression, and purification of P. berghei CSP
A fragment encoding aa Q24 – P92 of P. berghei CSP was amplified and fused to a fragment encoding aa P173–S326 of P. berghei CSP by an overlap extension PCR approach. The resulting fragment (encoding a CSP deletion mutant lacking the signal peptide, the N-terminal part of the repeat region, the region II and the GPI anchoring sequence but containing both the T cell epitope and the C-terminal part of the repeat region) was amplified using the primer pair 5′-ATAAGCTTCAAAATAAAAGCATCCAAGCCC-3′ / 5′-ATCTCGAGTCAACTTGAACATTTATCCATTTTAC-3′ and cloned into the prokaryotic expression vector pJC45 using the HindIII and XhoI restriction sites. The His10-PbCSP fusion protein (molecular weight 33 kDa) was recombinantly expressed in E. coli pAPLacI and purified by Ni-NTA affinity chromatography under native conditions.
ELISA plates (Millipore, MA) were coated overnight at 4°C with recombinant CSP (5 μg/mL) and blocked with 1% BSA afterwards. Mouse sera in serial dilutions from 1:50 to 1:6400 in PBS were incubated on the plates over night at 4°C. CSP-specific IgG was detected with alkaline phosphatase-conjugated anti-mouse pan-IgG (Promega, 1:6000 in PBS) and p-nitrophenylphosphate substrate (Sigma, Taufkirchen, Germany). OD was measured at 405 nm. Titers were defined as highest dilution that gives a positive signal above background, cut-off was calculated as mean OD of naïve serum plus three times standard deviation.
In vitro sporozoite inhibition assay
HepG2 cells were seeded at 1 × 105 per well into Nunc 8 well chamber slides (Thermo Fischer Scientific, Langenselbold, Germany) and grown over night in MEM with 10% FCS at 37°C. Prior to infection P. berghei sporozoites were preincubated 30 min on ice with naïve control serum or immune sera in 1:10 dilution in MEM with 3% BSA. Afterwards cells were infected with 10 000 sporozoites per well in MEM with 3% BSA. After 3 h the infection medium was removed and cells were grown for another 40 h in MEM with 10% FCS. For detection of P. berghei infected HepG2 cells the parasitophorous vacuole (PV) was stained. After fixation with 4% PFA and methanol the PV protein ExpI was stained with anti-ExpI (chicken) followed by a secondary anti-chicken Ig TRITC-conjugate (Jackson ImmunoResearch, Suffolk, UK) together with DAPI. The number of infected cells per well was counted under the microscope. The percent inhibition of sporozoite invasion was calculated in relation to the number of infected cells after infection with sporozoites in the absence of serum.
Statistical significance was analyzed using the Prism software (Graph Pad Software, San Diego, CA) as stated in the legends to the figures.
T.J. was funded by the collaborative research centre 841 (SFB 841). P.S. was supported by grant No. 310/08/0447 and the Research Plan AV0Z50200510.
Conflict of interest
The authors declare no financial or commercial conflict of interest.