2.1 Mouse-antisera generated after immunization with recombinant Pfs230 D1M domain show high reactivity in vitro
Pfs230 is a 230 kDa cysteine-rich protein, originally present as a 360 kDa precursor on the gametocyte surface(24). It includes 14 cysteine-rich domains (CM) and a natural protease cleavage site at position 542 (25). Previous studies have reported that high transmission-blocking activity can be achieved using the CM1 domain as an immunogenic antigen (23). In addition, analyses of polymorphisms within that region revealed only two predominant amino-acid substitutions at positions G605S and K661N, with the G605S having the highest frequency (AF 0.94) (23). A low polymorphism frequency in the targeting epitope is a desirable trait that reduces the risk of escaper mutations arising in the parasite that would impair the efficacy of the antibody. Twelve new putative missense mutations have been recently identified in the same region; however, they are based on de novo variant call data that require further validation (Fig.S3) (26).
Our selected antigen for BALB/c mice immunization comprised a 195 amino acids region from the cleavage site in position 542 through the end of the cysteine-rich domain 1, which we refer to as the Pfs230 D1M domain according to previous publications (23, 26) (Fig. 1A). Test bleeds were collected after the 3rd antigen boost to assess antibody titers elicited by immunization according to standard protocols. Indirect-ELISAs (enzyme-linked immunosorbent assays) with antiserum collected from each individual mouse were used to check the antibody titrations (as illustrated in Fig. 1B). All samples were found to be reactive at a 1:512,000 dilution with mice S4 and S5 showing the highest antibody titers (Fig. 2A), confirming the highly immunogenic properties of the Pfs230 D1M domain.
2.2 Mouse-antisera generated by immunization with the Pfs230 D1M domain significantly reduces oocyst loads
After assessing the immunogenic response elicited by the Pfs230 D1M domain antigen, we evaluated the anti-Plasmodium activities of all mouse-antisera through a standard membrane-feeding assay (SMFA). Immunized mice were boosted 3 times and 50 µL of antiserum from each mouse was kept after each boosting, pooled, and used to isolate the IgG fraction. The IgG fraction from each mouse was then supplied to Plasmodium falciparum gametocytes cultures blood mix (with RBC and human serum) to a final concentration of 250 µg/ml and fed to the Anopheles female mosquitoes through a membrane feeder (Fig. 1B). The IgG fraction isolated from pre-immunized mice was used as a negative control (Fig. 2B, Ctl-IgG), together with the group of mosquitoes fed on the gametocytes blood mix supplied with PBS as the mock control (Fig. 2B, Pf-only). Since the transmission-blocking activity of previously characterized Pfs230-specific antibodies was complement-dependent (27), the human serum in the blood meal was not heat-inactivated. The infectious blood meal was delivered with high gametocytaemia to achieve a strong infection prevalence and intensity that would facilitate the selection of the most effective anti-Plasmodium IgGs.
The in vitro reactivities from all immunized mice (S1-S5) were comparable (Fig. 2A), however, IgGs isolated from mouse S2 and S5 showed a significantly higher level of transmission-reducing activity (Fig. 2B). Similar to previous studies describing the anti-Plasmodium activity of Pfs230 (28), we found a prominent reduction in oocysts number in infected mosquito midguts (8-fold reduction of median oocyst load with S5 IgG, Mann-Whitney test, p < 0.0001; and a significant reduction of infection prevalence, Fisher’s exact test, p < 0.01) (Fig. 2B, Fig. 2C). Taking these results together, we selected mouse S5 for hybridoma production to produce monoclonal antibodies.
2.3 Hybridoma supernatant reacts to native Pfs230
We employed well-established hybridoma technology to produce monoclonal antibodies. Briefly, B lymphocytes isolated from immunized mice were fused with immortal myeloma cell lines to form the hybridoma (29). B-lymphocytes isolated from mouse S5 were used to generate 20 hybridoma cell lines (1E3, 1F11, 3D1, 3D6, 3F10, 3G11, 4B6, 4G8, 7A7, 9F3, 11A2, 12D9, 12E1, 12H6, 13G9, 14D2, 14F11, 15A3, 15F8, 15E9), each expressing a monoclonal antibody targeting the Pfs230 D1M domain. To validate whether hybridoma-produced monoclonals were interacting with the Pfs230 D1M domain, we performed in vitro indirect-ELISA (Fig. 1B). Secondary staining with a Fcy fragment-specific peroxidase-AffiniPure goat anti-mouse IgG revealed high immunoreactivities for all 20 undiluted hybridoma supernatants, with 450 nm OD readings ranging from 2.239 to 2.874 (Fig. 3A), thereby confirming that all 20 hybridomas produce monoclonal antibodies that can bind the Pfs230 D1M domain antigen.
To assess whether the antibodies could recognize the native Pfs230 protein on the surface of P. falciparum NF54 gametocytes, we performed immunofluorescence assays (IFAs) (Fig. 1B) with hybridoma supernatants. Staining with a secondary Fcγ fragment-specific Peroxidase-AffiniPure Goat Anti-Mouse IgG, showed that only the supernatants collected from clones 13G9, 3F10, and 14D2 bound strongly to gametocytes (Fig. 3B), and these were thus selected for further studies to assess their parasite-blocking potential. Clones 1E3, 1F11, 3D1, 3D6, 3G11, 4B6, 4G8, 7A7, 9F3, 11A2, 12D9, 12E1, 12H6, 14F11, 15A3, 15F8 and 15E9 displayed a reactivity comparable to the negative control with unnoticeable binding of any antibodies to the gametocytes (data not shown).
2.4 The 13G9 monoclonal antibody has a potent Plasmodium-blocking activity
Next, in order to assess the efficacy of candidate monoclonal antibodies to suppress P. falciparum infection, we first isolated IgG fractions from hybridoma’s supernatants 13G9, 3F10, 14D2 as testing groups, and IgG from complete hybridoma cell media as the negative control (Ctl-IgG) and evaluated their anti-Plasmodium activities by SMFA (Fig. 1B). Their reactivity to the gametocytes was again confirmed by the same assays as described above (Fig. 1B). Both in vitro indirect-ELISA and immunofluorescence have confirmed the specific activities of these three monoclonal antibodies (Fig. 4).
Co-feeding Anopheles females 30 µg of 13G9 monoclonal antibody with P. falciparum gametocytes-infected blood meal (with a final concentration of 166 µg/ml) through a membrane feeder resulted in a prominent reduction of infection intensity (Fig. 5A) (2.5-fold reduction of mean oocyst load, Mann-Whitney test p = 0.0035) and infection prevalence (50% reduction, Fisher’s exact test, p = 0.0011) (Fig. 5B) at 8 days post-infection. Monoclonal antibodies 3F10 and 14D2 did not show any Plasmodium-blocking activity despite exhibiting binding to both the Pfs230 D1M domain and the full-length protein in vitro (Fig. 4), highlighting the necessity for in vivo functional assays in addition to the in vitro reactivity assays.