T cell epitope regions of the P. falciparum MSP1-33 critically influence immune responses and in vitro efficacy of MSP1-42 vaccines

Abstract

The C-terminal 42 kDa fragments of the P. falciparum Merozoite Surface Protein 1, MSP1-42 is a leading malaria vaccine candidate. MSP1-33, the N-terminal processed fragment of MSP1-42, is rich in T cell epitopes and it is hypothesized that they enhance antibody response toward MSP1-19. Here, we gave in vivo evidence that T cell epitope regions of MSP1-33 provide functional help in inducing anti-MSP1-19 antibodies. Eleven truncated MSP1-33 segments were expressed in tandem with MSP1-19, and immunogenicity was evaluated in Swiss Webster mice and New Zealand White rabbits. Analyses of anti-MSP1-19 antibody responses revealed striking differences in these segments' helper function despite that they all possess T cell epitopes. Only a few fragments induced a generalized response (100%) in outbred mice. These were comparable to or surpassed the responses observed with the full length MSP1-42. In rabbits, only a subset of truncated antigens induced potent parasite growth inhibitory antibodies. Notably, two constructs were more efficacious than MSP1-42, with one containing only conserved T cell epitopes. Moreover, another T cell epitope region induced high titers of non-inhibitory antibodies and they interfered with the inhibitory activities of anti-MSP1-42 antibodies. In mice, this region also induced a skewed TH2 cellular response. This is the first demonstration that T cell epitope regions of MSP1-33 positively or negatively influenced antibody responses. Differential recognition of these regions by humans may play critical roles in vaccine induced and/or natural immunity to MSP1-42. This study provides the rational basis to re-engineer more efficacious MSP1-42 vaccines by selective inclusion and exclusion of MSP1-33 specific T cell epitopes.

Figure 1. Aligned amino acid sequences of the eleven truncated MSP1-42 subunit protein constructs compared to MSP1-42.

All constructs contain the MSP1-19 fragment (not shown) at the C-terminal end.

Figure 2. SDS-PAGE of the purified S2 cell expressed truncated MSP1-42 proteins.

Expected molecular sizes of each construct are in parenthesis. Lane 1: Construct 33-A (19 kDa); Lane 2: Construct 33-B (21 kDa); Lane 3: Construct 33-C (14 kDa); Lane 4: Construct 33-D (29 kDa); Lane 5: Construct 33-E (32 kDa); Lane 6: Construct 33-F (39 kDa); Lane 7: Construct 33-G (21 kDa); Lane 8: Construct 33-H (19 kDa); Lane 9: Construct 33-I (17 kDa); Lane 10: Construct 33-J (21 kDa); Lane 11: Construct 33-K (19 kDa).

Figure 3. Truncated MSP1-42 proteins possess disulfide sensitive conformation.

Immunoblots of recombinant proteins separated under reducing (lanes 1) and non-reducing (lanes 2) conditions and probed with conformational sensitive anti-MSP1-19 monoclonal antibodies. Panel A: Construct 33-A – 33-C and MSP1-19 probed with mAb 12.8; Panel B: Constructs 33-A – 33-C and MSP1-19 probed with mAb 2.2; and Panel C: Constructs 33-A – 33-C and MSP1-19 probed with mAb 5.2 , .

Figure 4. ELISA antibody responses against MSP1-19 in Swiss Webster mice immunized with recombinant truncated MSP1-42 proteins.

Panel A, percent responsiveness of mice immunized with Constructs 33-A – 33-K after the first booster injection (grey) and after the second booster injection (black). Panel B, antibody titers of mice vaccinated with Constructs 33-A – 33-K. Results of tertiary bleeds are shown. Horizontal bars indicate mean antibody titers. ANOVA (p<0.05) indicated that the levels of antibody titers differed among groups. Asterisk indicates a significant difference (Turkey post-hoc test, p<0.05) between Construct 33-D and all other vaccination groups.

Figure 5. Induction of MSP1-specific IL-4 (grey bars) and IFN-γ (white bars) responses, as determined by ELISPOT, in mice immunized with truncated MSP1-42 proteins.

Panel A: Constructs 33-A – 33-D; Panel B: Constructs 33-E – 33-K. Horizontal bars indicate mean SFU. Logistic Regression for Repeated Measures indicated that IFNγ levels were significantly higher (p<0.05) in Construct 33-E – 33-K compared to Construct 33-A – 33-D. No significant difference was found when comparing IL-4 levels. Mouse splenocytes were harvested 21 days after the last immunization.

Figure 6. Class II epitope prediction of the sequence of Construct 33-I by computer algorithm (Propred).

Grey shaded sequences represent motifs that may bind to Class II molecules.