Memory B-Cell and Antibody Responses Induced by Plasmodium falciparum Sporozoite Immunization

Background Immunization of healthy volunteers during receipt of chemoprophylaxis with Plasmodium falciparum sporozoites (CPS-immunization) induces sterile protection from malaria. Antibody responses have long been known to contribute to naturally acquired immunity against malaria, but their association with sterile protection after whole sporozoite immunization is not well established. We therefore studied the induction and kinetics of malaria parasite antigen-specific antibodies and memory B-cells (MBCs) during CPS-immunization and their correlation with protection from challenge infection. Methods We assessed humoral reactivity to 9 antigens representing different stages of the life cycle of P. falciparum by performing standardized MBC enzyme-linked immunospot and enzyme-linked immunosorbent assays on peripheral blood mononuclear cells and plasma samples from 38 Dutch volunteers enrolled in 2 randomized controlled clinical trials. Results MBCs and antibodies recognizing pre-erythrocytic and cross-stage antigens were gradually acquired during CPS-immunization. The magnitude of these humoral responses did not correlate with protection but directly reflected parasite exposure in CPS-immunization and challenge. Conclusions Humoral responses to the malarial antigens circumsporozoite protein, liver-stage antigen-1, apical membrane antigen-1, and merozoite surface protein-1 do not to predict protection from challenge infection but can be used as sensitive marker of recent parasite exposure. Clinical Trials Registration NCT01236612 and NCT01218893.

Fourteen CPS-immunized volunteers were assigned to receive a challenge with Pf-parasitized erythrocytes (blood-challenge BC; n=9) or a sporozoite challenge by infected mosquito bites (MC; n=5). Ten additional volunteers were assigned to two control groups for either BC (n=5) or MC (n=5) (Fig. 1A). CPSimmunization was conducted with three times 15 mosquitoes (eight mosquitoes infected with Pf strain NF54 and seven mosquitoes infected with Pf 3D7 (clone of NF54)). Twenty-one weeks following immunization (17 weeks after discontinuation of chloroquine prophylaxis), all volunteers were subjected to a challenge infection. MC was performed by exposure to bites of five 3D7-infected mosquitoes. BC was conducted by intravenous administration of 1962 viable 3D7 Pf-infected erythrocytes. All ten control subjects (MC and BC) and all nine immunized BC volunteers showed parasitized erythrocytes on thick smear and quantitative (q) PCR, whilst the five immunized MC volunteers remained thick smear negative until 21 days after challenge and were then presumptively drug-treated.

Study B [2]
Twenty-nine volunteers were randomly assigned to four groups, three CPS-immunization groups and one control group (Fig. 1B). CPS-immunization was carried out using three different numbers of NF54-infected mosquitoes: Study subjects were subjected three times to bites of 15 infected mosquitoes (3x15; n=5), ten infected (and five uninfected) mosquitoes (3x10; n=9) or five infected (and ten uninfected) mosquitoes (3x5; n=10). Control subjects received bites from 15 uninfected mosquitoes (n=5). Therefore, data from 29 challenged subjects were available for analysis. Nineteen weeks following immunization (15 weeks after discontinuation of chloroquine prophylaxis), all volunteers were subjected to a MC challenge infection by bites of five NF54-infected mosquitoes. All five control subjects developed blood-stage parasitemia, as detected by thick smear and qPCR. Of the 24 immunized subjects, sterile protection from MC challenge (defined as blood-stage parasite negative by thick smear until day 21 after challenge) was observed in 4/5 volunteers in the 3x15 group, 8/9 in the 3x10 group and 5/10 of the 3x5 group. The remaining seven CPSimmunized subjects were not protected, but did show a mean delay of 2.5 days in prepatent period of blood-stage parasitemia by both thick smear and qPCR compared to the control subjects. Twenty-one days after challenge, all volunteers remaining thick smear negative were presumptively drug-treated.

Sample collection
Sampling of citrate anti-coagulated peripheral blood for immunological analysis was performed at different time points in Study A and Study B using CPT vacutainers (Becton Dickinson). In Study A, samples were available from D0 (pre-immunization, before onset of chloroquine prophylaxis), day before challenge (C-1) 2 and 35 days after challenge (C+35). In Study B, additional sampling was performed one month (28 days) after each of the three immunizations ( Fig. 1).

ELISpot assay
MultiScreen Filter PVDF Immobilon plates (MSIPS4510, Millipore) pre-treated with 35% Ethanol were coated overnight at 4°C with 10µg/ml monoclonal antibodies to human IgG (clones MT91/145; Mabtech) or 4µg/ml of one of the malaria antigens (all dilutions in PBS). Plates were washed thoroughly and blocked with 1% bovine serum albumin (BSA; Sigma-Aldrich) in RPMI for 2h at 37°C. 4x10 5 mitogen-stimulated PBMCs per well were seeded in quadruplicates into the malaria antigen-coated filter plates. Anti-IgG coated wells were seeded with 1200 or 4000 cells/well. All time points for one volunteer were measured on the same plate. Filter plates were incubated for 6h at 37°C, 5% CO 2 . After washing, immobilized IgG antibody in the proximity of ASCs was detected using polyclonal goat anti-human IgG (Fc) alkaline phosphatase (1:1000 in PBST/0.5% FCS; Mabtech) overnight at 4°C. Plates were developed in the dark using alkaline phosphatase substrate BCIP/NBT (Mabtech), rinsed with water, left to dry and stored protected from light until reading using the CTL ImmunoSpot Reader (Cellular Technology Ltd.). 3

Background correction
Since pre-culture B-cell proportions (percentage of total PBMCs) and post-culture ASCs (per million PBMCs) were highly variable between donors (median with range 5.97 % (1.12-18.32 %) and 31,792

ELISA
Concentrations of malaria antigen-specific antibodies were determined in citrate anti-coagulated plasma in relation to a pool of 100 sera from adults living in a highly endemic area in Tanzania (HIT serum, [13]) by standardized enzyme-linked immunosorbent assay (ELISA). Polystyrene flat-bottom plates (NUNC™ Maxisorp, Thermo Scientific) were coated overnight at 4°C with 1µg/ml malaria antigen in PBS. After washing, plates were blocked with 3% BSA in PBS. Plasma samples were diluted in PBST/1% FCS and analyzed in duplicates. A four-point 1:2 dilution series was carried out for each sample. As a standard, duplicates of pooled HIT serum were included on every plate in a seven-point dilution series. The optimal dilution range was determined prior to the study for every antigen. All time points for one volunteer were measured on the same plate. Two-step detection was performed using biotinylated polyclonal goat antihuman IgG (Fc) (Mabtech; 1:1250) and streptavidin-conjugated horseradish peroxidase (HRP, Mabtech, 1:2000). All incubation steps from blocking to detection were carried out for 1h at 37°C in a humidified chamber. Plates were developed at room temperature using HRP substrate (Tetra-methyl-benzidine; tebubio); the reaction was stopped using 0.2M H 2 SO 4 . Spectrophotometrical absorbance at 450nm was measured using the Anthos 2001 ELISA plate reader.

Analysis
The standard curve of HIT serum was plotted on a logarithmic scale and fitted to a power trend line (R 2 > 0.99), optical density (OD) measurements for each sample (average of duplicates that were no more than 15% different) were converted to arbitrary units (AU) in relation to HIT serum. For each antigen,  c. Exposure to both pre-erythrocytic and blood stages; n=7 from Study B d. Exposure to blood-stages only; n=9 from Study A e. Differences between time points were analyzed by Wilcoxon matched-pairs signed rank test. Significant differences are indicated by asterices: * (p<0.05), ** (p<0.01), *** (p<0.001).