- Open Access
Analysis of the dose-dependent stage-specific in vitro efficacy of a multi-stage malaria vaccine candidate cocktail
Malaria Journalvolume 15, Article number: 279 (2016)
The high incidence and mortality rate of malaria remains a serious burden for many developing countries, and a vaccine that induces durable and highly effective immune responses is, therefore, desirable. An earlier analysis of the stage-specific in vitro efficacy of a malaria vaccine candidate cocktail (VAMAX) considered the general properties of complex multi-component, multi-stage combination vaccines in rabbit immunization experiments using a hyper-immunization protocol featuring six consecutive boosts and a strong, lipopolysaccharide-based adjuvant. This follow-up study investigates the effect of antigen dose on the in vitro efficacy of the malaria vaccine cocktail using a conventional vaccination scheme (one prime and two boosts) and a human-compatible adjuvant (Alhydrogel®).
IgG purified from rabbits immunized with 0.1, 1, 10 or 50 µg doses of the VAMAX vaccine candidate cocktail was analysed for total IgG and antigen-cocktail-specific titers. An increase in cocktail-specific titers was observed between 0.1 and 1 µg and between 10 and 50 µg, whereas no significant difference in titers was observed between 1 and 10 µg. Antigen component-specific antibody titers and stage-specific in vitro efficacy assays were performed with pooled IgG from animals immunized with 1 and 50 µg of the VAMAX cocktail. Here, the component-specific antibody levels showed clear dose dependency whereas the determined stage-specific in vitro IC50 values (as a correlate of efficacy) were only dependent on the titer amounts of stage-specific antibodies.
The stage-specific in vitro efficacy of the VAMAX cocktail strongly correlates with the corresponding antigen-specific titers, which for their part depend on the antigen dose, but there is no indication that the dose has an effect on the in vitro efficacy of the induced antibodies. A comparison of these results with those obtained in the previous hyper-immunization study (where higher levels of antigen-specific IgG were observed) suggests that there is a significant need to induce an immune response matching efficacy requirements, especially for a PfAMA1-based blood stage vaccine, by using higher doses, better adjuvants and/or better formulations.
Malaria remains a major challenge for the healthcare systems and economies of many developing countries, especially in sub-Saharan Africa  and an efficient vaccine would help to address the problems associated with constantly evolving drug resistance of the parasite, Plasmodium falciparum . Several strategies can be chosen in the context of malaria vaccine development. Vaccines can target “lower hanging fruits” such as the prevention or reduction of clinical manifestation, pregnancy-associated malaria, and malaria transmission, or they can aim for the “holy grail” of sustained strain-transcending sterile protection. While the GSK vaccine Mosquirix®, based on circumsporozoite protein (PfCSP) [3–5], exclusively targets the pre-erythrocytic stage of P. falciparum to prevent the establishment of the parasite within the liver, other approaches focus on blood-stage antigens that can be found on the surface of merozoites, to induce immune responses that block the invasion of red blood cells and thereby prevent or reduce clinical episodes. A vaccine that reduces the blood-stage parasite load may also reduce transmission. In addition to PfMSP1 and fragments thereof [6, 7], PfMSP2  and PfMSP3 [9, 10], the apical membrane antigen PfAMA1 [11, 12] has been proposed as a promising blood-stage vaccine candidate. This molecule is highly polymorphic, so allele-specific immune responses show moderate or even low efficacy against heterologous strains of P. falciparum. The allelic diversity of PfAMA1 has been addressed by the design of three artificial, diversity-covering variants (DiCo1-3)  or by combining three , four , six  or seven  different PfAMA1 alleles. In the field of transmission-blocking vaccines (TBVs), a zygote and ookinete surface antigen remains the leading candidate  and is undergoing clinical investigation . The authors of the current article previously designed and analysed the stage-specific in vitro efficacy of a malaria antigen cocktail  comprising three recombinant fusion proteins combining the three PfAMA1_DiCo variants each with Pfs25, and PfMSP1-19 (VAMAX1), PfCSP_TSR (VAMAX2) or PfCelTOS (VAMAX4). The latter work demonstrated the ability of the multi-component vaccine cocktail (VAMAX1, 2 and 4) featuring antigens from different stages of the P. falciparum life cycle to elicit parasite growth-inhibitory responses against the pre-erythrocytic stage, the blood stage and the sexual stage. However, the authors observed the proteolytic degradation of VAMAX4, leading to the loss of the C-terminal fusion partner PfCelTOS. Modified variants of VAMAX4 were, therefore, investigated, and in VAMAX6 PfCelTOS was replaced with a promising epitope (Q5A)  derived from the blood-stage antigen PfRH5 . In contrast to the immunization experiments with the initial cocktail (VAMAX1, 2 and 4) which featured a hyper-immunization regime (high 200 µg prime dose and five boosts of 100 µg), the work described herein investigated the dose-dependency of immune responses and stage-specific in vitro efficacy for the modified cocktail (VAMAX1, 2 and 6) using a human-compatible adjuvant (Alhydrogel®) and a one prime/two boosts vaccination regime. Additionally, the authors sought to confirm the observed correlation between antigen abundance within the improved VAMAX cocktail and the proportions of the corresponding antigen-specific antibodies found in the rabbit immune IgG fraction.
Construct design and cloning
The design and cloning of constructs VAMAXl and VAMAX2 and selection of recombinant Pichia pastoris clones was previously described . VAMAX6 comprising PfDiCo3, Pfs25 and the epitope of blood-stage antigen RH5-specific, parasite-inhibitory monoclonal antibody Q5A (amino acids 198–213, GeneID PF3D7_0424100) was obtained as a Pichia pastoris codon-optimized synthetic gene from GeneArt (Invitrogen, Carlsbad, CA) (Fig. 1a). The construct was inserted as previously described  into a Pichia pastoris expression vector containing the methanol inducible AOX1 promoter and terminator to control transgene expression. Cloning was confirmed by DNA sequencing. The constructs did not contain any potential N-glycosylation motif which occur in the natural sequences of PfAMA1, and Pfs25. For details on the used knock-out mutations refer to the following publications [23, 24].
Transformation and screening of Pichia pastoris
The transformation, cultivation and screening of Pichia pastoris strain CBS704 was carried out as previously described .
Fed-batch fermentation and purification of the antigens
The pre-cultures were prepared and the cultivations were carried out as previously described [20, 25] with minor changes. The number of fermentation phases was reduced to two, so the process consisted only of a batch and an immediate induction phase. For the latter phase, the temperature was lowered to 25 °C and the methanol concentration was kept constant at 0.25 % (v/v) by the use of an ALKOSENS probe combined with an ACETOMAT NII controller (Heinrich Frings GmbH & Co. KG, Bonn, Germany). During induction, the dissolved oxygen tension continuously dropped to 0 % as the stirrer speed reached a maximum of 600 rpm. When a total of 2.7 kg methanol was added, the pH was adjusted to 7.0 followed by the harvest and centrifugation of the broth (9000×g, 20 min, 4 °C). The culture supernatant was collected for immediate processing or for storage at–20 °C. Antigens were purified by immobilized metal ion affinity chromatography (IMAC) using Chelating Sepharose Fast Flow (GE Healthcare Life Sciences, Little Chalfont, UK) as a capture step (the N-terminal pro-peptide of PfAMA1 included in all three constructs binds to copper-charged chelating Sepharose Fast Flow resin) followed by buffer exchange using a HiPrep 26/10 column (GE Healthcare) and anion exchange (AEX) chromatography using Q Sepharose HP (GE Healthcare). Finally, samples were concentrated with a Centriprep YM-10 concentrator (Merck, Darmstadt, Germany) and polished by size exclusion chromatography (SEC) on a Sephacryl S-100 HR 16/60 column (GE Healthcare).
LDS-PAGE and immunoblot analysis
Purified VAMAX1, VAMAX2 and VAMAX6 were fractionated separately under non-reducing conditions on 4–12 % (w/v) polyacrylamide gradient gels (NuPage, Thermo Fisher Scientific, Waltham, MA, USA) and either stained with Coomassie Brilliant Blue (Fig. 1b) or transferred onto a nitrocellulose membrane (Whatman, GE Healthcare) for immunoblot analysis as previously described . After blocking with 5 % (w/v) skimmed milk dissolved in phosphate buffered saline (PBS), the VAMAX1, 2 and 6 proteins were probed with the Pfs25-specifc monoclonal antibody mAb4B7 (obtained by MR4) as well as the plant-derived chimeric variant PfAMA1-specific mAb4G2 (provided by Stefan Menzel, Fraunhofer IME, Aachen, Germany) (Fig. 1c). Additionally, mAb6.75, a PfCSP_TSR-specific murine monoclonal antibody generated by Christoph Kühn (Fraunhofer IME), was used specifically to detect the C-terminal PfCSP of VAMAX1 and the murine mAb5.2 (obtained by MR4) enabled the specific detection of the C-terminal PfMsp1_19 of VAMAX2. All primary antibodies were used at a concentration of 1 µg/ml (Fig. 1d). The secondary antibody was an alkaline phosphatase-labeled goat anti-mouse H + L or alkaline phosphatase-labeled goat anti-human Fc (both from Jackson Immunoresearch, West Grove, PA, USA).
Mass spectrometry (MS)
Proteins fixed in polyacrylamide gels were reduced, alkylated and digested with trypsin (Promega, Mannheim, Germany) as previously described . The resulting peptides were analysed by nanoHPLC (UltiMate 3000 HPLC system, LC Packing, Dionex, Idstein, Germany) coupled to an amaZon ETD MS ion trap spectrometer (Bruker Daltonics, Bremen, Germany) using ESI nano sprayer. MS/MS spectra were searched for peptides of interest using the local search engine Mascot Search v2.3.01 (Matrix Science Ltd, London, UK). Specific information on the MS measurements was previously reported .
Each purified VAMAX fusion protein (VAMAX 1, VAMAX 2 and VAMAX 6) was dialyzed against the formulation buffer (2.4 mM NaH2PO4, 2.6 mM K2HPO4, 0.125 mM Na2EDTA, 0.27 M D(-)mannitol, pH 6.8) (slightly modified from ). An equimolar VAMAX mixture was prepared and filter sterilized using a Pall Acrodisc 32 mm Supor filter (pore size 0.2 µm). The equimolar VAMAX mixture was adjusted to 200 µg/ml (50 µg vaccine dose), 40 µg/ml (10 µg vaccine dose), 4 µg/ml (1 µg vaccine dose) or 0.4 µg/ml (0.1 µg vaccine dose) with filter-sterilized formulation buffer and 1.7 ml of each VAMAX mixture was filled into clear glass injection vials (2 ml, 13-mm crimp neck) and lyophilized. The vials were sealed with rubber stoppers and aluminum crimp seals. One vial was used per immunization time point and dose. For the alum formulation, a pre-dilution of GMP-grade Alhydrogel® was prepared by mixing 3.2 ml Alhydrogel® with 6.8 ml saline under sterile conditions. The Alydrogel® pre-dilution was mixed by gentle agitation and allowed to equilibrate for 60 min at 4 °C. A lyophilized vial was reconstituted with 1.7 ml saline under sterile conditions and four aliquots of 280 µl were transferred to sterile 1.5-ml reaction tubes. Samples were stored at 4 °C, and 280 µl of the Alhydrogel® pre-dilution was added to each vial to obtain the final immunization dose, resulting in 0.8 mg Al3+ per 500 µL immunization dose. The adsorption of the VAMAX mixture to Alhydrogel® under formulation conditions was determined to be >95 %.
Rabbit immunization, titer determination
Rabbits were housed, immunized and sampled by Biogenes GmbH (Berlin, Germany), according to national animal welfare regulations. After equilibration of the vials for 60 min at 4 °C, each rabbit was immunized with an injection dose of 500 µl. For each dose (0.1, 1, 10 and 50 µg), four rabbits were immunized on days 0 (prime), 28 (boost 1) and 56 (boost 2). Serum samples were taken on days 0 (pre-immune) and 70 (final bleed). Titer determination was carried out as previously described by direct ELISA using the VAMAX antigen cocktail to coat the ELISA plates (100 ng/well) .
Total IgG was purified by conventional Protein A affinity chromatography using 40 ml immune serum (day 70) from rabbits immunized with the 1-µg dose and 20 ml immune serum from rabbits immunized with the 50-µg dose. Bound antibodies were eluted with 100 mM glycine (pH 3.0) and directly neutralized by adding 10 % (v/v) 1 M Tris–HCl (pH 8.0). The elution fractions were concentrated to 2.5 ml using VIVASPIN 15R centrifugal concentration devices with a molecular weight cut off of 30 kDa (Sartorius, Göttingen, Germany), and exchanged against 5 mM phosphate buffer (pH 7.5) using disposable PD10 columns (GE Healthcare). After lyophilization, the samples were reconstituted in 600 µl PBS and filter sterilized (pore size 0.2 µm). Antibody integrity and the total IgG concentration were determined by analytical gel filtration as previously described . The final IgG preparations were used for calibration-free concentration analysis (CFCA) and all subsequent in vitro inhibition assays.
Calibration-free concentration analysis
The antigen-specific antibody concentrations were measured in the purified antibody preparations by CFCA [30, 31] using a Biacore T200 instrument. The purified antigens comprised PfAMA1 variants DiCo 1–3 (provided by Stephan Hellwig, Fraunhofer IME, with kind permission from Bart Faber, BPRC, The Netherlands), PfCSP_TSR (in the context of CCT, a fusion protein containing PfCSP_TSR, PfCelTOS and PfTRAP_TSR ), Pfs25 (in the context of F0, a fusion protein containing Pfs25gk and Pfs230_C0, ) and PfMSP1_19 (in the context of a tetravalent PfMsp1-19 fusion protein, provided by Güven Edgü, Fraunhofer IME). The antigens were separately covalently coupled to CM5-S-Series sensor chips by standard EDC-NHS chemistry as previously reported . To match the assay specifications in terms of the initial binding rates at 5 µl/min, purified rabbit immune IgG was used at a concentration of 20–100 µg/ml as determined by test injections. Initial binding rates at 5, 30 and 100 µl/min were used to calculate antigen-specific IgG concentrations using the CFCA tool in the Biacore T200 evaluation software.
Sporozoite gliding motility assay
Plasmodium falciparum NF54 sporozoites were isolated and used in an adapted sporozoite gliding motility (SGM) assay . Each triplicate assay required 10,000 sporozoites per well in a 96-well glass bottom black plate. After 90 min incubation at 37 °C, 98 % relative humidity, 93 % (v/v) N2, 4 % (v/v) CO2 and 3 % (v/v) O2, gliding trails were washed and stained with biotinylated anti-CSP monoclonal antibody 3SP2 followed by AlexaFluor594-labeled streptavidin (Invitrogen). An automated high-content microscope (Leica) was used to capture nine images per well at 1000× magnification. Images were automatically processed with FIJI imaging software, as described elsewhere (Bolscher et al., pers. comm.).
Sporozoite invasion and liver stage development assay
The sporozoite invasion and liver stage development (SILSD) assay  was adapted as follows. Cryopreserved human hepatocytes (Tebu-Bio) were thawed and cultured in a 96-well clear bottom black plate (BD Falcon) for 2 days according to manufacturer’s protocol. Per well, 50,000 Pf NF54 sporozoites were pre-incubated with purified rabbit IgG samples for 30 min on ice, after which the mixture was transferred onto the hepatocytes. Samples were tested in triplicate and the medium was refreshed daily according to the manufacturer’s protocol. On day 6 after sporozoite invasion, hepatocytes were washed and fixed with 4 % (v/v) paraformaldehyde. Vehicle and 3SP2 control wells were stained with rabbit anti-PfHSP70, followed by AlexaFluor594-labeled goat anti-rabbit IgG and 4′,6-diamidino-2- phenylindole (DAPI). Vehicle and rabbit IgG samples were stained with a pool of mouse monoclonal antibodies directed against PfBip (binding protein, a marker of the endoplasmic reticulum), PfEXP-1 (exported antigen AG 5.1), PfHSP70 (heat shock protein 70) and PfMSP1 (merozoite-specific protein 1) followed by AlexaFluor488-labeled chicken anti-mouse IgG and DAPI. To determine the number of positively-stained (infected) hepatocytes, an automated high-content microscope (Leica) was used to capture 25 images per well at 100× magnification. Images were automatically processed with FUJI imaging software, as described elsewhere (Bolscher et al., pers. comm.).
In vitro growth inhibition assay
The ability of purified polyclonal rabbit IgGs to inhibit the growth of P. falciparum strains 3D7, 7G8 and V1-S (obtained through the MR4) was determined using growth inhibition assays (GIAs) as previously described . Reversal of growth inhibition assays were carried out as previously described .
Standard membrane feeding assay
The standard membrane feeding assay (SMFA)  was adapted as follows. The diluted rabbit IgG sample (36 μl) was mixed with 300 μl PfNF54:HSP70:GFP:luc gametocyte culture adjusted to a final haematocrit of 44.7 % red blood cells in 44.6 % human serum-containing active complement. The mixture was fed to Anopheles stephensi mosquitoes. After 8 days, the mosquitoes were frozen at –20 °C and the next day 24 mosquitoes per cage were analysed for luciferase expression as a measure for oocyst intensities . As controls, 24 uninfected mosquitoes were analysed to determine background luminescence levels.
The antibody titers determined by CFCA were averaged (n = 4) for each combination of antibody type (VAMAX mix titer, total IgG titer, domain-specific titers) and sample group [antigen dose, pre-immune serum (P) and immune serum (I)] and the normal distribution of the data in each of these sample sets was tested using the Kolmogorow–Smirnow test (a = 0.05) . The corrected standard deviation of each sample set was used to calculate its variance. Variances were compared using a pairwise F-test between sample groups for each antibody type (a = 0.05). If differences between the variances were not significant and the data points were normally distributed, a two-sided t test (a = 0.05) was used to identify significant differences between the antibody titers of sample set pairs. In case multiple group comparisons were made, Šídák’s multiple comparisons post hoc test was used following analysis of variance (ANOVA). For the VAMAX mix titers, the data were normalized by log-transformation for analysis. The association between the relative proportion of component-specific IgG and the relative proportion of molecular weight calculated for the components was analysed by calculating the Spearman’s rank correlation coefficient. The IC50-values were estimated by non-linear regression using the following model:
Ethics approval and consent to participate
The human cells used in this study were provided by Human HepCell (Paris, France). The company obtained the permission, issued by the French Ministry of Education and Research Ethical Commitee (Decision n° AC 2011-1308 de la Cellule Bioéthique de la Direction Générale pour la Recherche et l´Innovation du novembre 2011) to provide such samples for research use. The donors were informed by their surgeon that the tissues collected during the surgery could be used for research purposes only and that they get no financial compensation for this donation. For each donor, a consent of non-opposition is signed.
Rabbits were housed, immunized and sampled by Biogenes GmbH (Berlin, Germany), according to national animal welfare regulations. The animal facilities and protocols were reviewed and approved by: Landesamt für Landwirtschaft, Lebensmittelsicherheit und Fischerei MecklenburgVorpommern (LALLF M-V) (Approval No: 7221.3-2-030-13). To isolate the blood after immunization according to national regulations the animals were anesthetized using Ventranquil, stunned using a captive bolt device and exsanguinated by throat cut.
Production of the antigen cocktail
Recombinant Pichia pastoris cultures individually transformed with the three expression constructs (VAMAX 1, 2 and 6, see Fig. 1a) were used for fed-batch fermentation (15-L scale). After purification by IMAC and SEC, the purity and integrity of the recombinant proteins was analysed by LDS-PAGE followed by Coomassie staining, immunoblotting and analytical SEC. Monoclonal antibodies were available for the specific detection of each protein domain within the three VAMAX fusion proteins except the PfRH5 Q5A epitope. The proteins were obtained at high purity (>95 % determined by analytical SEC) the apparent molecular masses agreed with the calculated values of 87,592 Da for VAMAX 1 (measured 87,463 Da), 89,906 Da for VAMAX 2 (measured 90,312 Da) and 80,864 Da for VAMAX 6 (measured 80,813 Da) (Table 1), there was no indication of relevant amounts of degradation products, and the specific monoclonal antibodies detected the corresponding antigen domains (Fig. 1b–e). The yields after purification were 10, 13 and 4 µg/ml (calculated after purification and given as µg fusion protein/ml culture supernatant) for VAMAX 1, VAMAX 2 and VAMAX 6, respectively. MALDI-TOF–MS was used to determine the identity and integrity of the three recombinant proteins by analysing the tryptic fragments and calculating the total mass. In accordance with the immunoblot data, tryptic fragments representing each of the protein domains were identified. Together with the total mass analysis (Table 1), where approximately the expected molecular masses were observed for the three recombinant fusion proteins (−0.2 % for VAMAX 1, +0.4 % for VAMAX 2 and −0.1 % for VAMAX 6) these results confirmed the presence of fully intact fusion proteins.
Immunization, determination of titers and total IgG levels
Immune sera from rabbits immunized with the different doses of the VAMAX 1, 2 and 6 cocktail were used for the determination of cocktail-specific titers. Figure 2a shows the titers of individual animals as well as the geometric mean obtained after immunization with the corresponding doses. Comparing the geometric means of the four animals per group, a greater than tenfold increase in titer (from 3.8 × 104 to 5.6 × 105) was observed between the 0.1 and 1 µg doses, which was statistically significant. Only a marginal difference in titer was observed between the 1 and 10 µg antigen doses (from 5.6 to 6.0 × 105) but the titers showed another statistically significant increase between the 10 and 50 µg doses (from 6.0 × 105 to 2.5 × 106). The observed titers led to the selection of samples from the 1 and 50 µg doses for pooling (hereafter P1 and P50, respectively) and subsequent analysis in the stage-specific parasite inhibition assays. Additionally, the total IgG in the sera of the animals before (P) and after immunization (I) were compared. As shown in Fig. 2b, the total IgG levels after immunization were independent of the antigen dose because there were no statistically significant differences between the geometric mean IgG levels between immune sera of the different dose groups, which all contained total IgG concentrations of ~14 mg/ml.
Determination antigen-specific IgG levels
CFCA was used to determine the amounts of antigen-specific IgG within pools of purified total IgG from immune sera in the P1 and P50 dose groups, enabling the correlation of inhibition values obtained from stage-specific in vitro assays not only with total IgG but also with antigen and/or stage-specific IgG concentrations (Table 2). The highest concentration of antigen-specific IgG was measured for the PfAMA1_DiCo mix, followed by Pfs25, PfMSP1-19 and PfCSP_TSR. Q5A-specific IgG could not be quantified since no Q5A peptide-specific reactivity could be detected in rabbit immune IgG by ELISA. The values (µg/ml antigen-specific IgG) were used to calculate percent values for the specificity of each antigen in relation to total IgG (Fig. 3). Here, pronounced differences (up to three-fold for the PfAMA1–DiCo mix) were observed between the P1 and P50 groups for all four antigens. The observed differences between the P1 and P50 groups were statistically significant with the exception of PfMSP1_19, although even in this case there was a slight tendency towards higher concentrations of PfMSP1_19-specific antibodies in the P50 group.
In vitro efficacy assays
The in vitro efficacy of the different doses of the VAMAX vaccine cocktail were compared using appropriate stage-specific in vitro parasite inhibition assays: pre-erythrocytic/liver stage (SGM and SILSD assays), blood stage (GIAs), and sexual stage (SMFAs). The assays were carried out using purified and pooled immune IgG preparations from the P1 and P50 dose groups.
Liver stage efficacy
The SGM assay was carried out at concentrations of 9, 3 and 1 mg/ml total IgG. Inhibition was measured as a percentage relative to the concentration of neutral rabbit serum and was plotted against total IgG (Fig. 4a) and the corresponding PfCSP-specific IgG (Fig. 4b). Because the inhibition values did not exceed 50 % even at the highest IgG concentration, the IC50 values were estimated for total IgG as well as PfCSP-specific IgG and summarized in Table 3. An inhibition value of ±50 % was observed for the highest total IgG concentration in the case of the P50 sample, whereas only ±20 % inhibition was observed at the same total IgG concentration for the P1 sample, reflecting the quantitative differences in PfCSP-specific IgG between the two dose groups (Fig. 3).
Additionally, the SILSD assay was carried out by combining sporozoites with 9 mg/ml total IgG onto cryopreserved primary human liver cells. In agreement with the SGM assay, different degrees of inhibition were observed (±40 % for P50 and ±30 % for P1), which again correlated with the quantitative differences in PfCSP-specific antibodies in the total IgG preparations of P1 and P50 (Fig. 4c).
Blood stage efficacy
Strain-dependent blood-stage-specific efficacy was addressed by carrying out GIAs using different P. falciparum strains (3D7, 7G8 and V1-S) to account for the major polymorphisms covered by the three DiCo variants. Figure 5a shows the strain-specific dose-dependency of growth inhibition for total immune IgG pools P1 and P50. The different IC50 values are summarized in Table 3. Additionally, a reversal of growth inhibition assay was carried out to investigate the contribution of PfAMA1_DiCo mix-specific IgG to blood stage-specific in vitro efficacy, using the 3D7 reference strain and PfAMA1_DiCo mix as a competitor. As shown in Fig. 5b the growth-inhibitory effect of rabbit immune IgG was completely neutralized when a mixture of PfAMA1_DiCo was used at equal or higher molar concentrations, compared to rabbit immune IgG antigen-binding sites. Because this finding indicates that the parasite growth-inhibitory activity measured in this assay is exclusively mediated by PfAMA1_DiCo mix-specific antibodies, it was possible to estimate IC50 values in relation to the concentrations of PfAMA1_DiCo mix-specific antibodies (Fig. 5c). The calculated IC50 values in relation to PfAMA1_DiCo mix-specific antibodies as well as total IgG are shown in Table 3. No significant differences of IC50 values were observed for the PfAMA1_DiCo mix-specific antibodies (120–180 µg/ml) or total IgG (8–30 mg/ml) between the P50 and P1 dose pools as well as between the different parasite strains.
Sexual stage efficacy
The potential transmission-blocking activity of the VAMAX 1, 2 and 6 specific rabbit immune IgG was analysed by SMFA. The transmission-blocking activity of sera derived from the P1 and P50 dose pools correlated with the concentration of Pfs25-specific IgG (Fig. 6). As shown in Table 3, the estimated IC50 did not deviate significantly between the P1 and P50 dose pools (1.409 and 1.049 µg/ml, respectively, for Pfs25-specific IgG).
After the predominantly disappointing results achieved with many single-component malaria vaccine candidates in clinical trials, the concept of combining different P. falciparum antigens to improve, immunogenicity, efficacy and strain coverage has received greater attention [38, 39]. The authors of this article are following different approaches towards the development of a multi-stage multi-component malaria vaccine cocktail. In this context, a vaccine cocktail was developed  comprising three recombinant fusion proteins based on the three diversity-covering (DiCo) variants of the P. falciparum apical membrane antigen PfAMA1 developed by Remarque et al. , each featuring Pfs25 fused to the C-terminus of the PfAMA1_DiCo variant, and additionally the TSR-domain of PfCSP (VAMAX 1), the 19 kDa C-terminal fragment of PfMSP1 (VAMAX 2) or PfCelTOS (VAMAX 4). A further approach based on a cocktail of VAMAX 1, 2 and 4 used a hyper-immunization protocol to induce high immune IgG titers, to generally confirm and dissect the multi-stage in vitro efficacy of the fusion antigen cocktail. In the current study, the dose-dependency of antigen-specific titers and in vitro efficacy was investigated based on a more realistic immunization scheme and a human-compatible adjuvant. Because proteolytic cleavage of the C-terminal PfCelTOS domain in the DiCo3-based VAMAX4 was previously observed, a modified variant (VAMAX 6) was used in this study, in which PfCelTOS was replaced with the Q5A epitope from the blood-stage antigen PfRH5.
All three fusion proteins (VAMAX 1, 2 and 6) were produced by Pichia pastoris fed-batch fermentation and a purity of >95 % was achieved after IMAC and SEC for all three proteins. The observed yields (VAMAX 1 10 µg/ml, VAMAX 2 13 µg/ml, VAMAX 6 4 µg/ml, calculated after purification given as µg fusion protein/ml culture supernatant) were within the previously observed range for the production and purification of the VAMAX 1, 2 and four variants  but lower than the yields originally reported for PfAMA1 (120 µg/ml, after purification) . Lower expression levels may result either from intrinsic features of the fusion partner, or from the increased size of the fusion protein for different PfAMA1_DiCo fusion proteins where yield was inversely related to the protein size . The integrity and conformation of the recombinant proteins was investigated by LDS-PAGE, immunoblot with antigen-domain-specific antibodies, and mass spectrometry, confirming the presence of intact, full-size proteins. There was no indication of proteolytic cleavage, previously reported for the VAMAX four component in which most of the C-terminal PfCelTos domain was found to be cleaved off from the fusion protein .
A one prime/two boost immunization scheme (protein in adjuvant formulation) of groups of four rabbits with four different doses (0.1, 1, 10 and 50 µg) of an equimolar mixture of the three recombinant fusion proteins (VAMAX 1, 2 and 6), using the human-compatible adjuvant Alhydrogel®, yielded a low titer (3.8 × 104) for the 0.1-µg dose, and reasonable titers (5.6 × 105–2.5 × 106) for the three higher doses with no differences between the 1 and 10 µg doses. This agrees with the results of a recent mouse immunization study using 0.01, 0.03, 0.1, 0.3 and 1 µg doses of PfAMA1_DiCo mix formulated in alum (Rehydragel®) which also showed clear dose-dependent seroconversion (Faber et al., pers. comm..). In contrast, no significant differences in total IgG levels were observed for the four different doses, indicating that the induced total IgG is adjuvant-dependent rather than antigen-dose-dependent in the investigated context.
Specific IgG concentrations were determined for all included antigen domains except the Q5A epitope, where no reactivity could be shown in ELISA. Even though Faber et al. have demonstrated that fusing smaller antigen domains like PfMSP1_19 to PfAMA1_DiCo could improve immune responses against a smaller domain that suffers from low immunogenicity when used alone , Q5A might be too small to sufficiently benefit from this effect in the context of the VAMAX6 fusion protein.
For the blood stage it was possible to attribute the growth inhibition activity (measured by GIA) exclusively to PfAMA1-specific antibodies because the addition of an excess of PfAMA1_DiCo mix completely abolished parasite growth inhibition in a reversal of growth inhibition assay. In this context, neither the PfMSP1_19 specific antibodies (P1 = 14 µg/ml and P50 = 43 µg/ml PfMSP1-19 in 10 mg/ml total IgG, respectively) nor any potential PfRH5_Q5A-specific antibodies (not determined) contributed to the inhibition of growth. As previously suggested , this phenomenon may reflect the presence of PfMSP1_19-specific and PfRH5_Q5A-specific antibodies that are below the concentration required for the reliable determination of growth inhibition in GIAs (>100 µg/ml for PfMSP1_19-specific antibodies  and >30 µg/ml for PfRH5_Q5A-specific antibodies ).
To investigate whether the in vitro protective efficacy of the induced immune responses is directly proportional to the amount of relevant, component-specific IgG, CFCA analysis was carried out using the different components (PfCSP_TSR, PfAMA1_DiCo mix, PfMSP1_19 and Pfs25). The estimated in vitro IC50 values observed for the pre-erythrocytic-stage antigen PfCSP_TSR (12 µg/ml for P50), the blood-stage antigen PfAMA1_DiCo mix (120–140 µg/ml for P50) and the sexual-stage antigen Pfs25 (1.0 µg/ml for P50) were in the same range as reported in other studies featuring or including these antigens in comparable assays [13, 20, 29, 42].
The correlation of these antibody concentrations with the estimated IC50 values observed in the corresponding stage-specific parasite-inhibition assays confirmed the absence of qualitative differences in terms of in vitro efficacy between antigen-specific IgG induced by different doses of the antigen cocktail, and that the efficacy in relation to total IgG is only dose-dependent in this context. Comparable IC50 values were also obtained for hyperimmunization with the predecessor cocktail (VAMAX 1, 2 and 4) using a strong lipopolysaccharide-based formulation (Biogenes proprietary adjuvant) , which further indicates that maximizing antigen-specific titers should be the main focus of further optimization efforts, preferentially with higher doses >50 µg, which should not be critical, when looking at the antigen doses used within approved complex combination vaccines such as Infanrix hexa® (GSK). Additionally, the potential of other human compatible adjuvants like Motanide ISA720 and AS02 (as shown for an PfAMA1-FVO-based vaccine candidate by Roestenberg et al.) or combinations thereof to significantly increase antibody responses against the VAMAX-mixture could be investigated . Furthermore, alternative particulate presentation formats like virus-like-particle fusions, currently being evaluated in a clinical trial (ClinicalTrials.Gov Id: NCT02013687) with the P. falciparum sexual stage antigen Pfs25, would open an additional strategy to improve immunogenicity.
The antigen component-specific IgG amounts showed an almost linear correlation with the quantitative representation of the corresponding component within the context of the mixture (Fig. 7), agreeing with the results presented in the previous study , and further emphasizing the potential to explore this correlation for the “fine tuning” of stage-specific functionality according to the strongly-biased IC50 requirements for parasite growth inhibition (at least based on in vitro efficacy in rabbits) in the context of malaria vaccine development. Small domains (e.g. PfMSP1-19) and promising inhibitory epitopes (e.g. PfRH_5Q5A) are most likely to require multivalent presentation, when used in combination with more and/larger antigens (e.g. PfAMA1).
This study demonstrates that the amounts of vaccine-specific antibodies induced following the immunization of rabbits with four different doses (0.1, 1, 10 and 50 µg) strongly depend on antigen dose when using a human-compatible adjuvant (Alhydrogel®) and vaccination scheme (one prime/two boosts). A comparison of the results with those obtained in an earlier study based on a hyper-immunization regime, in which higher levels of antigen-specific IgG were observed, suggests that there is significant need for improvement to generate a higher immune response and match efficacy requirements, especially for a PfAMA1-based blood stage vaccine, by using higher doses, better adjuvants and/or better formulations. The results also confirm the previously described strong positive correlation between antigen-specific antibody titers and the quantitative representation of the corresponding component within the vaccine mixture, suggesting an overrepresentation of smaller antigens or peptides e.g. by increasing valency or by formation of VLPs would be beneficial.
- IC50 :
half maximal inhibitory concentration
Plasmodium falciparum apical membrane antigen 1
Plasmodium falciparum circumsporozoite protein
Plasmodium falciparum merozoite surface protein
alcohol oxidase 1
immobilized metal ion affinity chromatography
size exclusion chromatography
polyacrylamide gel electrophoresis
high performance liquid chromatography
good manufacturing practice
enzyme-linked immunosorbent assay
phosphate buffered saline
calibration-free concentration analysis
standard membrane feeding assay
red blood cell
sporozoite gliding motility
sporozoite invasion and liver stage development
matrix-assisted laser desorption ionization
time of flight
WHO. World malaria report 2014. Geneva: World Health Organization; 2014.
Fairhurst RM, Nayyar GM, Breman JG, Hallett R, Vennerstrom JL, Duong S, et al. Artemisinin-resistant malaria: research challenges, opportunities, and public health implications. Am J Trop Med Hyg. 2012;87:231–41.
Wilby KJ, Lau TT, Gilchrist SE, Ensom MH. Mosquirix (RTS, S): a novel vaccine for the prevention of Plasmodium falciparum malaria. Ann Pharmacother. 2012;46:384–93.
Cohen J, Nussenzweig V, Nussenzweig R, Vekemans J, Leach A. From the circumsporozoite protein to the RTS, S/AS candidate vaccine. Hum Vaccin. 2010;6:90–6.
Tinto H, D’Alessandro U, Sorgho H, Valea I, Tahita MC, Kabore W, et al. Efficacy and safety of RTS, S/AS01 malaria vaccine with or without a booster dose in infants and children in Africa: final results of a phase 3, individually randomised, controlled trial. Lancet. 2015;386:31–45.
Pan WQ, Ravot E, Tolle R, Frank R, Mosbach R, Turbachova I, et al. Vaccine candidate MSP-1 from Plasmodium falciparum: a redesigned 4917 bp polynucleotide enables synthesis and isolation of full-length protein from Escherichia coli and mammalian cells. Nucleic Acids Res. 1999;27:1094–103.
Hui GS, Gosnell WL, Case SE, Hashiro C, Nikaido C, Hashimoto A, et al. Immunogenicity of the C-terminal 19-kDa fragment of the Plasmodium falciparum merozoite surface protein 1 (MSP1), YMSP1(19) expressed in S. cerevisiae. J Immunol. 1994;153:2544–53.
Conway DJ, Greenwood BM, McBride JS. Longitudinal study of Plasmodium falciparum polymorphic antigens in a malaria-endemic population. Infect Immun. 1992;60:1122–7.
Oeuvray C, Bouharountayoun H, Grasmasse H, Bottius E, Kaidoh T, Aikawa M, et al. Merozoite surface protein-3-a malaria protein inducing antibodies that promote Plasmodium-falciparum killing by cooperation with blood monocytes. Blood. 1994;84:1594–602.
Singh S, Soe S, Weisman S, Barnwell JW, Perignon JL, Druilhe P. A conserved multi-gene family induces cross-reactive antibodies effective in defense against Plasmodium falciparum. PLoS One. 2009;4:e5410.
Narum DL, Thomas AW. Differential localization of full-length and processed forms of PF83/AMA-1 an apical membrane antigen of Plasmodium falciparum merozoites. Mol Biochem Parasitol. 1994;67:59–68.
Remarque EJ, Faber BW, Kocken CH, Thomas AW. Apical membrane antigen 1: a malaria vaccine candidate in review. Trends Parasitol. 2008;24:74–84.
Remarque EJ, Faber BW, Kocken CHM, Thomas AW. A diversity-covering approach to immunization with Plasmodium falciparum apical membrane antigen 1 induces broader allelic recognition and growth inhibition responses in rabbits. Infect Immun. 2008;76:2660–70.
Kusi KA, Faber BW, Thomas AW, Remarque EJ. Humoral immune response to mixed PfAMA1 alleles; multivalent PfAMA1 vaccines induce broad specificity. PLoS One. 2009;4:e8110.
Dutta S, Dlugosz LS, Drew DR, Ge X, Ababacar D, Rovira YI, et al. Overcoming antigenic diversity by enhancing the immunogenicity of conserved epitopes on the malaria vaccine candidate apical membrane antigen-1. PLoS Pathog. 2013;9:e1003840.
Miura K, Herrera R, Diouf A, Zhou H, Mu J, Hu Z, et al. Overcoming allelic specificity by immunization with five allelic forms of Plasmodium falciparum apical membrane antigen 1. Infect Immun. 2013;81:1491–501.
Kusi KA, Faber BW, Riasat V, Thomas AW, Kocken CH, Remarque EJ. Generation of humoral immune responses to multi-allele PfAMA1 vaccines; effect of adjuvant and number of component alleles on the breadth of response. PLoS One. 2010;5:e15391.
Kaslow DC, Isaacs SN, Quakyi IA, Gwadz RW, Moss B, Keister DB. Induction of Plasmodium falciparum transmission-blocking antibodies by recombinant vaccinia virus. Science. 1991;252:1310–3.
Jones RM, Chichester JA, Manceva S, Gibbs SK, Musiychuk K, Shamloul M, et al. A novel plant-produced Pfs25 fusion subunit vaccine induces long-lasting transmission blocking antibody responses. Human Vaccin Immunother. 2015;11:124–32.
Spiegel H, Boes A, Kastilan R, Kapelski S, Edgue G, Beiss V, et al. The stage-specific in vitro efficacy of a malaria antigen cocktail provides valuable insights into the development of effective multi-stage vaccines. Biotechnol J. 2015;10:1651–9.
Douglas AD, Williams AR, Knuepfer E, Illingworth JJ, Furze JM, Crosnier C, et al. Neutralization of Plasmodium falciparum merozoites by antibodies against PfRH5. J Immunol. 2014;192:245–58.
Baum J, Chen L, Healer J, Lopaticki S, Boyle M, Triglia T, et al. Reticulocyte-binding protein homologue 5-An essential adhesin involved in invasion of human erythrocytes by Plasmodium falciparum. Int J Parasitol. 2009;39:371–80.
Kocken CH, Withers-Martinez C, Dubbeld MA, van der Wel A, Hackett F, Valderrama A, et al. High-level expression of the malaria blood-stage vaccine candidate Plasmodium falciparum apical membrane antigen 1 and induction of antibodies that inhibit erythrocyte invasion. Infect Immun. 2002;70:4471–6.
Beiss V, Spiegel H, Boes A, Kapelski S, Scheuermayer M, Edgue G, et al. Heat-precipitation allows the efficient purification of a functional plant-derived malaria transmission-blocking vaccine candidate fusion protein. Biotechnol Bioeng. 2015;112:1297–305.
Spiegel H, Schinkel H, Kastilan R, Dahm P, Boes A, Scheuermayer M, et al. Optimization of a multi-stage, multi-subunit malaria vaccine candidate for the production in Pichia pastoris by the identification and removal of protease cleavage sites. Biotechnol Bioeng. 2015;112:659–67.
Boes A, Spiegel H, Edgue G, Kapelski S, Scheuermayer M, Fendel R, et al. Detailed functional characterization of glycosylated and nonglycosylated variants of malaria vaccine candidate PfAMA1 produced in Nicotiana benthamiana and analysis of growth inhibitory responses in rabbits. Plant Biotechnol J. 2015;13:222–34.
Shevchenko A, Tomas H, Havlis J, Olsen JV, Mann M. In-gel digestion for mass spectrometric characterization of proteins and proteomes. Nat Protoc. 2006;1:2856–60.
Roestenberg M, Remarque E, de Jonge E, Hermsen R, Blythman H, Leroy O, et al. Safety and immunogenicity of a recombinant Plasmodium falciparum AMA1 malaria vaccine adjuvanted with Alhydrogel (TM), Montanide ISA 720 or AS02. PLoS One. 2008;3:e3960.
Boes A, Spiegel H, Voepel N, Edgue G, Beiss V, Kapelski S, et al. Analysis of a multi-component multi-stage malaria vaccine candidate-tackling the cocktail challenge. PLoS One. 2015;10:e0131456.
Pol E, Karlsson R, Roos H, Jansson A, Xu B, Larsson A, et al. Biosensor-based characterization of serum antibodies during development of an anti-IgE immunotherapeutic against allergy and asthma. J Mol Recognit. 2007;20:22–31.
Williams AR, Douglas AD, Miura K, Illingworth JJ, Choudhary P, Murungi LM, et al. Enhancing blockade of Plasmodium falciparum erythrocyte invasion: assessing combinations of antibodies against PfRH5 and other merozoite antigens. PLoS Pathog. 2012;8:e1002991.
Voepel N, Boes A, Edgue G, Beiss V, Kapelski S, Reimann A, et al. Malaria vaccine candidate antigen targeting the pre-erythrocytic stage of Plasmodium falciparum produced at high level in plants. Biotechnol J. 2014;9:1435–45.
Behet MC, Foquet L, van Gemert GJ, Bijker EM, Meuleman P, Leroux-Roels G, et al. Sporozoite immunization of human volunteers under chemoprophylaxis induces functional antibodies against pre-erythrocytic stages of Plasmodium falciparum. Malar J. 2014;13:136.
Maskus DJ, Bethke S, Seidel M, Kapelski S, Addai-Mensah O, Boes A, et al. Isolation, production and characterization of fully human monoclonal antibodies directed to Plasmodium falciparum MSP10. Malar J. 2015;14:276.
Roeffen W, Theisen M, van de Vegte-Bolmer M, van Gemert G, Arens T, Andersen G, et al. Transmission- blocking activity of antibodies to Plasmodium falciparum GLURP.10C chimeric protein formulated in different adjuvants. Malar J. 2015;14:443.
Vos MW, Stone WJR, Koolen KM, van Gemert GJ, van Schaijk B, Leroy D, et al. A semi-automated luminescence based standard membrane feeding assay identifies novel small molecules that inhibit transmission of malaria parasites by mosquitoes. Sci Rep. 2015;5:18704.
Rees DG. Essential Statistics. 4th ed. Oxford: Taylor & Francis; 2000.
Halbroth BR, Draper SJ. Recent developments in malaria vaccinology. Adv Parasitol. 2015;88:1–49.
Hill AVS, Biswas S, Draper S, Rampling T, Reyes-Sandoval A. Towards a multi-antigen multi-stage malaria vaccine. Malar J. 2014;13(Suppl 1):O31. doi:10.1186/1475-2875-13-S1-O31.
Faber BW, Remarque EJ, Kocken CH, Cheront P, Cingolani D, Xhonneux F, et al. Production, quality control, stability and pharmacotoxicity of cGMP-produced Plasmodium falciparum AMA1 FVO strain ectodomain expressed in Pichia pastoris. Vaccine. 2008;26:6143–50.
Faber BW, Younis S, Remarque EJ, Rodriguez Garcia R, Riasat V, Walraven V, et al. Diversity covering AMA1-MSP119 fusion proteins as malaria vaccines. Infect Immun. 2013;81:1479–90.
Cheru L, Wu YM, Diouf A, Moretz SE, Muratova OV, Song GH, et al. The IC50 of anti-Pfs25 antibody in membrane-feeding assay varies among species. Vaccine. 2010;28:4423–9.
AB and HS conceived the study, performed the experiments, analysed the data and wrote the manuscript. NV generated Pichia pastoris expressions constructs and clones and contributed to writing the manuscript. RK performed Pichia pastoris fermentation and purification of the antigens and contributed to writing the manuscript. IC performed analytical mass spectrometry and contributed to data analysis and writing the manuscript. JB and KD developed and performed the SGM, SILSD and SFMA assays and contributed to data analysis and writing the manuscript. RF and SB performed GIA assays and contributed to data analysis and writing the manuscript. JFB performed statistical analysis and contributed to writing the manuscript. AR, SS and RFi conceived the overall study design and contributed to writing the manuscript. All authors read and approved the final manuscript.
We thank Güven Edgü (Fraunhofer IME) for providing recombinant PfMSP1_19. We thank Stephan Hellwig (Fraunhofer IME) for providing purified recombinant PfAMA1DiCo1-3 with the kind permission from Ed Remarque and Bart Faber (BPRC Rijswijk, The Netherlands) who also generously provided us with PfAMA1-specific polyclonal rabbit IgG (BG98). The following reagents were obtained through the MR4 as part of the BEI Resources Repository (NIAID, NIH, Bethesda, MD, USA): Mab4B7, MRA-28 and Mab5.2, MRA94 deposited by D.C. Kaslow, Plasmodium falciparum 3D7A, MRA-151, deposited by D. Walliker and P. falciparum 7G8, MRA-152, deposited by D. Walliker, P. falciparum V1-S, MRA-176 deposited by D.E. Kyle. We also thank Prof. Dr. Jean Paul Noben and his colleagues from Laboratory of Residue Analysis, Biomedical Research Institute, Limburgs Universitair Centrum in Diepenbeek, Belgium, for the MS analysis of intact fusion proteins.
AB, HS, AR, and RFi have filed a patent on multi stage malaria vaccines (EP14183995.1, US-provisional US62/047,286). All authors declare that they have no competing interests.
This work was supported by the “Fraunhofer Zukunftsstiftung” and the Frauhofer-Gesellschaft Internal Programs under Grant No. Attract 125-600164.
Alexander Boes and Holger Spiegel contributed equally to this work