Cellular and antibody response in GMZ2-vaccinated Gabonese volunteers in a controlled human malaria infection trial

Abstract

Background: Antibody and cellular memory responses following vaccination are important measures of immunogenicity. These immune markers were quantified in the framework of a vaccine trial investigating the malaria vaccine candidate GMZ2.

Methods: Fifty Gabonese adults were vaccinated with two formulations (aluminum Alhydrogel and CAF01) of GMZ2 or a control vaccine (Verorab). Vaccine efficacy was assessed using controlled human malaria infection (CHMI) by direct venous inoculation of 3200 live Plasmodium falciparum sporozoites (PfSPZ Challenge). GMZ2-stimulated T and specific B-cell responses were estimated by flow cytometry before and after vaccination. Additionally, the antibody response against 212 P. falciparum antigens was estimated before CHMI by protein microarray.

Results: Frequencies of pro- and anti-inflammatory CD4⁺ T cells stimulated with the vaccine antigen GMZ2 as well as B cell profiles did not change after vaccination. IL-10-producing CD4⁺ T cells and CD20⁺ IgG⁺ B cells were increased post-vaccination regardless of the intervention, thus could not be specifically attributed to any malaria vaccine regimen. In contrast, GMZ2-specific antibody response increased after the vaccination, but was not correlated to protection. Antibody responses to several P. falciparum blood and liver stage antigens (MSP1, MSP4, MSP8, PfEMP1, STARP) as well as the breadth of the malaria-specific antibody response were significantly higher in protected study participants.

Conclusions: In lifelong malaria exposed adults, the main marker of protection against CHMI is a broad antibody pattern recognizing multiple stages of the plasmodial life cycle. Despite vaccination with GMZ2 using a novel formulation, expansion of the GMZ2-stimulated T cells or the GMZ2-specific B cell response was limited, and the vaccine response could not be identified as a marker of protection against malaria. Trial registration PACTR; PACTR201503001038304; Registered 17 February 2015; https://pactr.samrc.ac.za/TrialDisplay.aspx?TrialID=1038.

Keywords: CHMI; Cytokine; GMZ2; Memory B cells; Microarray; P. falciparum.

Fig. 1

GMZ2-CAF01 study overview. Participants were allocated in five separate groups: (1) Group A (control vaccine; n = 8, (2) Group B (100 µg GMZ2-Alhydrogel; n = 12), (3) Group C (30 µg GMZ2- CAF01; n = 8), (4) Group D (100 µg GMZ2-CAF01; n = 12), and (5) Group E (100 µg GMZ2-CAF01, without subsequent CHMI; n = 10). All injections were administered intramuscularly in the deltoid muscle on study days 0, 28, and 56, in alternating sides. Thirteen weeks after receiving last completion of the immunization dose, volunteers of Groups A–D underwent CHMI by direct venous inoculation (DVI) of 3200 aseptic, purified, cryopreserved P. falciparum sporozoites (Sanaria® PfSPZ Challenge), strain NF54, to assess vaccine efficacy (VE). Group E volunteers were followed-up for 6 months post–immunization without CHMI. Small graph in the upper side corner indicates sampling time points selected to characterize T, B, and antibody breadth responses. Figure legend to the right indicates the time points designed to sample study subjects

Fig. 2

CD4⁺ T cell frequencies following immunization. Isolated PBMCs were stimulated with either medium alone, the vaccine antigen GMZ2, or Staphylococcal enterotoxin B (SEB) as positive control. Thereafter, intracellular cytokine staining was performed, and the cells measured by flow cytometry. Data are expressed after subtraction of unstimulated cell frequencies from that of stimulated with the positive control (SEB), and with GMZ2, and normalization with the average of positive control values. The comparison of the pro-inflammatory cytokine producing CD4⁺ T cells (Fig. 2a), and the anti-inflammatory cytokine producing CD4⁺ T cells (Fig. 2b) between D0 and D84 was performed in those receiving the control vaccine and in those vaccinated with GMZ2 (including those vaccinated with 100 µg GMZ2-Alhydrogel (opened dots), 30 µg GMZ2-CAF01 (grey dots), 100 µg GMZ2-CAF01 (black dots) using Wilcoxon test with Bonferroni correction for multiple comparisons. p value less than 0.05 is considered as statistically significant. All time points per volunteer were measured in a single experiment after several optimization tests, and individual volunteers were measured in separate experiments. Symbols represent individual samples. Red lines represent the median values with interquartile range.

Fig. 3

GMZ2-specific B cell frequencies following immunization. B cells were estimated using cryopreserved PBMCs without additional stimulation. The comparison of the GMZ2-specific memory B cells frequencies between D0 and D84 was performed in those receiving the control vaccine and in those vaccinated with GMZ2 (including those vaccinated with 100 µg GMZ2-Alhydrogel (opened dots), 30 µg GMZ2-CAF01 (grey dots), 100 µg GMZ2-CAF01 (black dots) using Wilcoxon test with Bonferroni correction for multiple comparisons. p value less than 0.05 is considered as statistically significant. All time points per volunteer were measured in a single experiment after several optimization tests, and individual volunteers were measured in separate experiments. Symbols represent individual samples. Red lines represent median values with interquartile range

Fig. 4

Association between T and B phenotypes with the trial outcome. Dot plot graphs show the relation between pre- and post-immunization GMZ2-specific immune phenotypes regarding presence/absence clinical malaria status after CHMI. Any parasitaemia with symptoms or parasitaemia above 1000 parasites per µl was defined as malaria (black spots), whereas any parasitaemia below the threshold of 1000 parasites / µl with no symptoms (Control) as well as individuals with neither parasitaemia nor symptoms (Protected) are represented by open circles. Comparison of cytokine producing CD4⁺ T-, CD20 + B- and the GMZ2-specific B phenotypes was performed using Mann-Whitney (T cells) or unpaired t-tests (B cells). Data are from a single experiment after several optimization tests, and individual volunteers were measured in separate experiments. Symbols represent individual samples. Red lines represent median values and interquartile range. p value lower than 0.05 is considered significant

Fig. 5

Time to treatment regarding natural acquired immunity before vaccine intervention. Graphs show the time to first malaria treatment regarding the fraction of specific triple and double positive CD4⁺ T, CD20⁺IgG⁺ GMZ2^± and GMZ2-specific-CD27^± B cells (a), or the estimated cell number for each of the same cell phenotypes (b) at baseline (D0). Values above the median are represented in red whereas data below the median are shown in blue. The Log-rank test was used to compare the two curves. P value lower than 0.05 is considered significant

Fig. 6

Plasma IgG concentrations against GMZ2. IgG antibody concentration was measured by enzyme-linked immunosorbent assay (ELISA) in all study volunteers against the vaccine immunogen GMZ2 (a) as well as the GLURP (b) and MSP3 (c) fragments. Plasma was collected at days 0 and 84. Every violin plot represents the distribution within each interventional group. Bold lines indicate the median, and dotted lines mean the quartiles. The Wilcoxon matched pairs signed rank test were used to assess vaccine immunogenicity between days 0, and 84. p value lower than 0.05 is considered significant

Fig. 7

Correlation between GMZ2^−/+ B cell phenotypes and the anti-GMZ2 IgG concentration at baseline. The association between GMZ no specific CD20⁺IgG⁺ or CD20⁺IgG⁺GMZ2⁺ B cells and the anti-GMZ2 IgG concentration was performed on D0 data using Pearson’s correlation after log transformation. Data were obtained from separate experiments after several optimization tests. Symbols represent individual samples. p value lower than 0.05 is considered significant

Fig. 8

Protein microarray using plasma from volunteers undergoing CHMI. Figure shows the heatmap from protein microarrays in the semi-immune study population as well as a European malaria-naïve control population (a). The intensity of antibody responses in the population being protected from clinical malaria and those who developed malaria were compared and shown as volcano plot. The red circles are antigens being at least two-fold higher and significantly upregulated in the respective group (b). c shows the heatmap in participants having at least two-fold higher antibody response in the protected group vs. the unprotected group, thus showing the raw data of the red dots in Fig. 8b. Gene ID according to PlasmoDB are given. d shows the breadth of the antibody response in those who develop malaria (n = 15), those who control the parasitaemia (n = 12), those having full protection (n = 7), in the negative control (1 European malaria-naïve sample, measured in 5 technical replicates) and in European naïve controls (n = 13). Data are from a single experiment after several optimization tests, and individual volunteers were measured in separate experiments. Differential antibody recognition in the different allocated study outcomes was analysed by Student’s t-test. p value less than 0.05 is considered as statistically significant. NC Negative control, N malaria-naïve subjects, M subjects having clinical symptoms of malaria after CHMI, C subjects controlling parasitaemia, P subjects fully protected after CHMI