Preclinical Assessment of Viral Vectored and Protein Vaccines Targeting the Duffy-Binding Protein Region II of Plasmodium Vivax

Abstract

Malaria vaccine development has largely focused on Plasmodium falciparum; however, a reawakening to the importance of Plasmodium vivax has spurred efforts to develop vaccines against this difficult to treat and at times severe form of relapsing malaria, which constitutes a significant proportion of human malaria cases worldwide. The almost complete dependence of P. vivax red blood cell invasion on the interaction of the P. vivax Duffy-binding protein region II (PvDBP_RII) with the human Duffy antigen receptor for chemokines (DARC) makes this antigen an attractive vaccine candidate against blood-stage P. vivax. Here, we generated both preclinical and clinically compatible adenoviral and poxviral vectored vaccine candidates expressing the Salvador I allele of PvDBP_RII - including human adenovirus serotype 5 (HAdV5), chimpanzee adenovirus serotype 63 (ChAd63), and modified vaccinia virus Ankara (MVA) vectors. We report on the antibody and T cell immunogenicity of these vaccines in mice or rabbits, either used alone in a viral vectored prime-boost regime or in "mixed-modality" adenovirus prime - protein-in--adjuvant boost regimes (using a recombinant PvDBP_RII protein antigen formulated in Montanide(®)ISA720 or Abisco(®)100 adjuvants). Antibodies induced by these regimes were found to bind to native parasite antigen from P. vivax infected Thai patients and were capable of inhibiting the binding of PvDBP_RII to its receptor DARC using an in vitro binding inhibition assay. In recent years, recombinant ChAd63 and MVA vectors have been quickly translated into human clinical trials for numerous antigens from P. falciparum as well as a growing number of other pathogens. The vectors reported here are immunogenic in small animals, elicit antibodies against PvDBP_RII, and have recently entered clinical trials, which will provide the first assessment of the safety and immunogenicity of the PvDBP_RII antigen in humans.

Keywords: Duffy-binding protein; MVA; Plasmodium vivax; adenovirus; blood-stage; malaria; poxvirus; vaccine.

Figure 1

Humoral responses induced by viral vectored and protein vaccines targeting PvDBP_RII. (A) BALB/c mice (n = 5) were immunized with 10¹⁰ vp of a recombinant HAdV5 expressing PvDBP_RII, and boosted 8 weeks later with 10⁷ pfu of a recombinant MVA expressing PvDBP_RII and the selectable marker GFP. (B) BALB/c mice (n = 6/group) were immunized with 1.5 × 10⁸ ifu of either HAdV5 or ChAd63 expressing PvDBP_RII, and 8 weeks later, mice were boosted with 10⁷ pfu MVA-PvDBP_RII either expressing the selectable marker GFP or no marker (ML). (C) BALB/c mice (n = 4–6/group) were immunized with the regime outlined in (A) (AM), or immunized three times i.m. 3 weeks apart with 10 μg of PvDBP_RII protein formulated with the specified adjuvant (PPP), or primed i.m. with 10¹⁰ vp of HAdV5-PvDBP_RII and boosted 8 weeks later i.m. with 10 μg of PvDBP_RII protein formulated with adjuvant (AP). In all panels, serum IgG titers were measured in arbitrary units (AU) against PvDBP_RII protein by ELISA 2 weeks after each immunization (day 14 and 70) and before the boost (day 55) in (A,B), and 2 weeks after the second immunization (PP) or final immunization (PPP, AP, AM) in (C). In (C), responses following two immunizations are shown with open symbols (PP, AP, AM) and after three immunizations with closed symbols (PPP). ND, not done. Median and individual data points are shown. The dotted line indicates the threshold for responses above background in (B) determined using serum taken from the mice prior to any immunization. The same cut-off would apply in (A) but is not indicated.

Figure 2

IgG isotype responses induced by PvDBP_RII vaccination. BALB/c mice (n = 4–6/group) were immunized with the AM, AP, or PPP regimes as described in Figure 1C. IgG1 and IgG2a antibody responses were measured in the serum 2 weeks after the last immunization by ELISA against PvDBP_RII protein. (A) Individual and median responses are shown. The dotted line indicates the threshold for responses above background determined using serum taken from the mice prior to any immunization. (B) IgG1:IgG2a ratios were calculated and log transformed for the AP and PPP groups. Individual and median responses are shown.

Figure 3

T cell responses induced by viral vectored and protein vaccines targeting PvDBP_RII. (A) BALB/c mice (n = 4–5/group) were immunized with AP and PPP regimes against PvDBP_RII as described in Figure 1C. Ten weeks after the last immunization, spleens were harvested and T cell responses were measured from frozen spleen samples by ex vivo IFN-γ ELISpot following re-stimulation with 5 μg/mL recombinant PvDBP_RII protein (rDBP). Median and individual data points are shown. (B) BALB/c mice (n = 6) were immunized with HAdV5-PvDBP_RII and boosted 8 weeks later with MVA-PvDBP_RII using doses as in Figure 1B. Two weeks after the last immunization, splenic T cell responses were measured by ex vivo IFN-γ ELISpot following re-stimulation with 5 μg/mL rDBP or 2 μg/mL OLP. Median and individual responses are shown. (C) Splenic T cell responses were measured from frozen samples in the mice reported in Figure 1B 2 weeks after the boost immunization and using OLP. Median and individual responses are shown. (D) BALB/c mice were immunized with 1.5 × 10⁸ ifu HAdV5-PvDBP_RII and 8 weeks later were boosted with 10⁷ pfu MVA-PvDBP_RII (GFP). Splenic T cell responses were measured from frozen spleen samples harvested 2 weeks post-boost and following re-stimulation with 5 μg/mL individual peptides (1–32) or the OLP pool control. Results from a representative mouse are shown. (E) BALB/c mice were immunized with 1 × 10⁸ ifu HAdV5-PvDBP_RII and 8 weeks later were boosted with 10⁷ pfu MVA-PvDBP_RII. Spleens were harvested 15 days post-boost. Intracellular cytokine staining followed by flow cytometric analysis showed the IFN-γ responses induced to the OLP pool and the three strongest peptides (Table 1) were from CD4⁺ cells. Cells were gated on lymphocytes, and then (top row) singlets by forward scatter height (FSC-H) versus area (FSC-A), then live cells (dead cell marker versus side scatter area, SSC-A), and then CD4 versus CD8. Representative plots (bottom row) of IFN-γ versus SSC-A from CD4⁺ gated cells are shown from one mouse following no stimulation (Unstim) or re-stimulation with the OLP pool or peptide 14 (P14).

Figure 4

Inhibition of PvDBP_RII binding to its receptor by vaccine-induced antibodies. BALB/c mice (n = 4–5/group) were immunized with the AM, AP, and PPP regimes as described in Figure 1C. Binding inhibition assays were carried out as described in Section “Materials and Methods” using serum from the stated time-points at dilutions of 1:1000, 1:2000, and 1:4000. Each sample was run in duplicate or triplicate and the mean% binding inhibition calculated. The mean and SD of the% binding inhibition for the (A) AM, (B) PPP, and (C) AP groups are shown. The controls (n = 3) were high-titer mouse anti-PvDBP_RII serum samples. (D) Median and individual data points are shown for the 1:4000 serum dilution reported in (A–C). Data for day 55 (D55) after HAdV5 are also included. *P < 0.05 and **P < 0.01, see Results for statistical tests. (E) Relationship between% binding inhibition and log₁₀ anti-PvDBP_RII IgG titer assessed by endpoint ELISA. Individual data points and non-linear regression curve are shown (n = 87).

Figure 5

Indirect IFA using serum from adenovirus-MVA PvDBP_RII immunized mice and rabbits. (A) Indirect IFA using sera from PvDBP_RII and OVA control immunized mice (green) and microscope slides containing fixed P. vivax-infected cells obtained from patients in Thailand. Representative images are shown for both sets of sera. Nuclei were stained with DAPI. The merge plus bright field are also shown. (B) ZiKa rabbits (n = 4) were immunized with HAdV5-PvDBP_RII on day 0 and MVA-PvDBP_RII on day 56. Serum was harvested before immunization (day 0) and 28, 56, and 70 days post-adenovirus administration. Serum IgG titers were determined by endpoint ELISA. Individual responses are shown and the solid line indicates the mean. The dotted line indicates the cut-off for a positive response – samples tested at 1:100 dilution that did not show an OD₄₀₅ nm reading above negative control sera are plotted below this line. Rabbits immunized with the same vectors encoding OVA showed no detectable responses in the same ELISA assay (not shown). (C) Indirect IFA as in (A) using sera from the PvDBP_RII and OVA immunized rabbits (green). Two representative images are shown for both sets of sera.