Novel sporozoite-based ELISpot assay to assess frequency of parasite-specific B cells after vaccination with irradiated sporozoites

Plasmodium falciparum 3D7 strain was obtained from the insectary at the Walter Reed Army Institute of Research. Recombinant, full length P. falciparum CSP (3D7) was produced in Escherichia coli as previously described [15].

Sporozoites (SPZ) were prepared by dissecting mosquitoes 16–20 days post blood feed using the Ozaki method [16], or the clean dissection method. The Ozaki method is a fast and simple dissection of mosquito heads followed by centrifugation of the material through a glass wool pillow; for the clean dissection method [17], salivary glands are removed from the mosquito heads, pooled and then disrupted by shear forces using a syringe and 18 g needle. SPZ were frozen as pellets and stored at − 80 °C to be used as lysates. Upon thawing, pellets were resuspended in PBS (pH 7.2), vigorously pipetted using siliconized/pre-lubricated pipette tips (to minimize loss of material), and lysis confirmed by microscopy.

Peripheral blood mononuclear cells (PBMC) from previously conducted clinical trials [IMRAS; Immunization by mosquito bite with radiation attenuated SPZ (NCT01994525) and MAL027 (NCT00075049)] were used. In the IMRAS trial, study participants received 5 doses of approximately 200 infectious bites/dose by P. falciparum-infected Anopheles stephensi mosquitoes. The goal of the trial was, apart from safety and immunogenicity, to develop reagents for the identification and validation of biomarkers of protection (Epstein et al. pers.commun.). In the MAL27 trial, study participants were immunized with three doses of RTS,S adjuvanted with AS01B or AS02A [18]. Cells from the MAL27 trial were used to represent ‘CSP-immune’ individuals. PBMC were thawed and counted using a Luna-FL™ Dual Fluorescence cell counter (fluorescence protocol with AO/PI to determine cell viability). Warm culture medium [(RPMI-1640 containing 10% fetal bovine serum (FBS), Pen/Strep, l-glutamine, non-essential amino acids (NEAA), Sodium Pyruvate, 2-mercaptoethanol] was prepared. Cell numbers were adjusted to 1.5 × 10⁶ cells/ml and stimulated with 1 μg/ml of the TLR7/8 ligand R848 (Mabtech) and 10 ng/ml recombinant human (rh) IL-2 (Mabtech) in culture medium for 36 to 48 h at 37 °C in a humidified incubator with 5% CO₂ in six-well plates (Corning, Tewksbury, MA) Stimulating B cells polyclonally with R848 and supporting full activation with recombinant IL-2 ensures that all B cell clones are activated and produce immunoglobulins. Sporozoite material on the ELISpot plates enabled the capturing of malaria-specific antibodies secreted while incubating the cells in the ELISpot plates.

PVDF-based membrane plates (MSIP, Millipore, Burlington, MA, USA) were activated with 15 µl/well 35% ethanol for 1 min following the manufacturer’s instructions. The plates were washed five times with tissue culture grade, sterile water. Next, the activated ELISpot plates were coated with 100 µl/well that equate to 30,000 lysed SPZ/well or recombinant CSP protein (1 µg/ml). The optimal amount of SPZ was determined in a previous study [11]. Assay controls were: (a) Bovine Serum Albumin (BSA, cell culture grade, Sigma Aldrich, St. Louis, MO, USA, 1 µg/ml) as the negative control to determine non-specific binding of antibodies to plates; (b) IgM-specific monoclonal antibody (mAb) MT11/12 (Mabtech) or IgG specific mAb MT91/145 (both at 15 µg/ml; Mabtech) as positive control since they capture any secreted antibody regardless of specificity. Plates were incubated overnight at 4 °C [11]. Liquid from wells coated with sporozoite lysate was gently removed and wells allowed to air dry since immobilizing the sporozoite material by air drying rather than fixation is superior regarding the reactivity with antibodies [11]. On the day of the experiment, plates were blocked with culture medium (RPMI 1640 with 10% FBS and supplements) for 2 h at 37 °C. It was necessary to use FBS in the culture medium since human AB pooled serum caused significant background issues. Next, R848-activated PBMC or CD19 enriched B cells were plated onto the ELISpot plates. The cell concentration of PBMC was 5 × 10⁵ cells/well (the maximum cell number recommended by the manufacturer of the ELISpot reagents) while the cell concentration of enriched B cells was 5 × 10⁴ to 10⁵. Cells were plated in triplicates. For B cell enrichment, R848-activated PBMC were used as starting material for the magnetic enrichment of B cells using CD19-microbeads (Miltenyi Biotec, Auburn, CA, USA) following the manufacturer’s instructions. In brief, polyclonally activated PBMC were harvested and cell numbers determined using a Luna-FL™ cell counter. Cells were pelleted (centrifugation at 300g for 10 min) and resuspended in degassed MACS buffer [PBS (pH 7.2) with 0.5% BSA and 2 mM EDTA] at a concentration of 10⁷ cells/100 µl. Twenty µl of anti-CD19 microbead-conjugated mAb was added and incubated for 15 min at 4 °C. Two ml of MACS buffer was then added and cells spun down to remove unbound antibody (300 g, 10 min). Finally, cells were resuspended in 500 µl MACS buffer and passed through MS columns (Miltenyi Biotec) inserted into an OctoMACS magnet (Miltenyi Biotec). Enriched cells were counted (Luna-FL™) and plated at a cell concentration of 5 × 10⁴ to 10⁵ cells/well. Plates were incubated for 24 h at 37 °C in a humidified incubator with 5% CO₂ to allow for the capturing of sporozoite-specific antibodies. Cells were removed from the plates by washing (Biotek EL405 plate washer, Biotek Winooski, VT, USA) plates five times with 250 µl/well PBS. Detection of spots was achieved by adding biotinylated mAb MT22 (for IgM) or mAb MT78/145 (for IgG) at 1 µg/ml (100 µl /well; PBS with 0.5% FBS as diluent) for 2 h at room temperature. Plates were washed and incubated with alkaline phosphatase-conjugated Streptavidin–alkaline (1:1,000, PBS with 0.5% FBS as diluent). All reagents were obtained from Mabtech. After 1 h at room temperature, plates were washed and BCIP/NBT substrate added until spots were visible. Reaction was stopped by washing plates with tap water. Dried ELISpot plates were analysed using the AID Autoimmun Diagnostika GmbH ELISpot reader (Strassberg, Germany) and accompanying software.

Statistically significant differences between the various assay conditions were determined by using two-sided T tests (Minitab 17, State College, PA, USA).

A recent report described optimal assay conditions for a sporozoite-based ELISA [11]. In that study, monoclonal antibodies as well as sera from RAS-immune subjects recognized both, sporozoite lysates and intact SPZ. The results from that study demonstrated that, for convenience, sporozoite lysates can be used in lieu of freshly isolated SPZ. In the present study, ELISpot plates were coated under the same conditions with sporozoite lysate or with recombinant CSP (as positive control) or BSA (negative control). To determine the impact of the sporozoite preparation, SPZ were prepared by two different methods: (1) crude preparation using the Ozaki method [16]; (2) clean preparation by dissecting salivary glands from the mosquito heads. For the ease of performing the experiments and based on previous data obtained from the sporozoite-based ELISA, SPZ were isolated using both preparation techniques and then frozen for later use. PBMCs from four CSP-immune donors were plated at a cell concentration of 5 × 10⁵ cells/well and tested for reactivity (Fig. 1). The clean dissection provided stronger signals for all four donors. Therefore, all subsequent experiments were conducted using SPZ from cleanly dissected salivary glands. It is not clear why the crude sporozoite preparation yielded lower responses, but one possible explanation is that debris interferes with the reactivity of the antibodies.

Initial experiments revealed that the signal strength was significantly dependent on the vaccine or malaria-exposure of the donors. To determine the detection threshold of the assay and to also increase the sensitivity of the assay, the following two experiments were conducted: (1) titre the cell number of CSP-immune PBMC to determine the dose response and detection threshold (Fig. 2); (2) compare the signal strength when using either CSP-immune PBMC or CSP-immune B cells in the assay (Fig. 3).

Titrating PBMC revealed that a linear response was achieved with cell numbers of between 10⁵ to 5 × 10⁵ cells/well. Adding more cells (10⁶ cells/well) did not increase signal strength. In contrast, plating a higher cell number than 5 × 10⁵ cells/well is not recommended by the manufacturer. A range of 2 × 10⁵ to 5 × 10⁵ is optimal and provides flexibility in cases where the sample volume is limited.

Human peripheral blood consists of 1% to 7% B cells. This raises the question whether purifying B cells with magnetically labelled CD19 could increase the sensitivity of the assay because it would be equivalent to plating a larger number of PBMCs (plating 10⁵ B cells would equate to 1.4–10 × 10⁶ PBMC), which is not possible due to crowding of cells in the well. Comparative experiments were set up whereby PBMCs versus positively enriched B cells (using CD19 microbeads) from each donor were tested for reactivity in the sporozoite-based ELISpot. Three volunteers from a previously conducted malaria clinical trial (RTS,S vaccine) were tested to assess parasite responses for pre-immune, post-3 (pre-challenge; 2 weeks after the last immunization) and post challenge (4 weeks after controlled human malaria infection) time points. Purified B cells showed higher spots per well as compared to unfractionated PBMCs (Fig. 3). The results demonstrated a statistically significant increase in the signal between purified B cells and PBMCs. Therefore, subsequent assays were always conducted with purified B cells after the polyclonal stimulation of PBMCs.

The next step after establishing the primary conditions for the SPZ-based ELISpot assay was to determine the frequency of IgM and IgG producing B cells that recognize epitopes on P. falciparum SPZ. Pre-immune, post 3 (i.e., pre-challenge; 2 weeks post last immunization), and post challenge (4 weeks post controlled human malaria infection) time points from two different CSP-immune donors were tested for reactivity in the SPZ-based ELISpot. After the overnight incubation of purified B cells in the ELISpot plates, cells were removed and the plates either probed with biotinylated anti-human IgM or anti-human IgG monoclonal antibodies followed by Streptavidin-AP to assess the frequency of CSP-specific IgM or IgG producing B cells (Fig. 4). The results demonstrate that vaccination results in significant increases in the number of SPZ-specific IgM and IgG-producing B cells. A weak response was observed in pre-immune samples; this could reflect cross-reactivities of pre-existing antibodies to other pathogens such as Toxoplasma. Similarly, there is drop in the frequency of SPZ-specific B cells after a controlled human malaria infection (challenge). This could be due to the sequestration of SPZ-specific B cells in immunological organs and warrants further analysis.

The next step was to apply the SPZ-based ELISpot assay to human vaccine models by testing pre-immune and pre-challenge PBMCs from a RAS clinical trial under the established conditions (10⁵ purified B cells/well). Results were stratified based on the protective status of the donors (n = 8 for protected subjects and n = 4 for not-protected subjects); in the IMRAS trial, donors were immunized by receiving approx. 960 infectious bites of P. falciparum infected, irradiated An. stephensi mosquitoes and protective efficacy assessed in a controlled human malaria infection (CHMI). In the CHMI study, participants are challenged with five bites of P. falciparum-infected mosquitoes and evaluated daily for the presence of blood-stage parasites. Only donors that remain parasite-free after 28 days are considered protected (i.e., sterile protection). IgG responses were measured against SPZ and CSP (Fig. 5). The results provided several important observations: (a) the frequency of CSP-specific B cells is significantly lower compared to B cells reactive to SPZ. While the data confirm that CSP is one of the major antigens recognized by malaria-immune B cells, they also show that other sporozoite-specific antigens are present and recognized by a significantly higher number of B cells; (b) the frequency of CSP-specific B cells is significantly higher in subjects protected by the RAS vaccine (p = 0.016, 2-sided T-test); (c) the frequency of sporozoite-specific B cells has a strong trend to be higher in protected subjects (p = 0.08, 2-sided T-test). The latter observation suggests the utility of this readout for identifying immune correlates of protection. The role of CSP-specific antibodies, and therefore B cells, in mediating sterile protection has long been recognized [11, 15, 19, 20]. Not all antigens recognized by IMRAS-induced antibodies contribute to protection. The weaker correlation between frequency of sporozoite-specific B cells and protection (Fig. 5b) would suggest that some of the recognized antigens either do not correlate with protection or may even correlate with susceptibility.