Immunization of Aotus Monkeys with Recombinant Plasmodium falciparum Hybrid Proteins Does Not Reproducibly Result in Protection from Malaria Infection

Abstract

Plasmodium falciparum antigens SERP, HRPII, MSAI, and 41-3 have shown promise as vaccine components. This study aimed at reproducing and extending previous results using three hybrid molecules. Antibody responses were reproduced in Aotus monkeys, but solid protection from a P. falciparum blood-stage challenge that showed an unintendedly enhanced pathogenicity was not observed.

The increasing drug resistance of Plasmodium falciparum, the most pathogenic human malaria parasite, underlines the need for an effective malaria vaccine. Identification, testing, and optimization of candidate molecules originating from all developmental stages of the parasite are under way. Previously, a successful trial in Aotus monkeys employed the Escherichia coli-expressed hybrid proteins MS2/SERP/HRPII and SERP/MSAI/HRPII (11). Both hybrid proteins contain a region of the serine repeat protein SERP (1, 9), including two putative T-cell epitopes (13) and previously shown to induce a partial protective response in Aotus monkeys (5), and the C-terminal half of the histidine-rich protein HRPII (14), which has also been shown to induce a partially protective response (5, 8). SERP/MSAI/HRPII contains in addition a conserved N-terminal region of the merozoite surface antigen MSAI (7) that includes at least four T-cell epitopes (3, 6). Here we report on further analysis of three hybrid proteins of this type in a vaccination trial with Aotus monkeys. Two of the proteins, SERP/HRPII and SERP/MSAI/HRPII, are improved versions of the hybrid proteins mentioned above, obtained by deleting nonmalaria protein regions and changing an internal restart residue (methionine-729 of SERP) into alanine. Thus, the SERP/HRPII hybrid protein comprises residues 630 to 893 of SERP fused to the 189 C-terminal residues of HRPII, and SERP/MSAI/ HRPII comprises residues 630 to 764 of SERP fused to residues 146 to 259 of MSAI, which is fused to the 189 C-terminal residues of HRPII. SERP/41-3/HRPII contains the same components as SERP/HRPII, and additionally includes residues 77 to 188 of antigen 41-3 (10), which was previously shown to confer protection against a P. falciparum challenge (5). The internal restart residue (methionine-100) was also mutagenized into alanine and another residue, arginine-319, was changed into leucine to prevent proteolytic degradation. SERP/MSAI/HRPII was partially purified to a final purity of about 30%, as described previously (8), in order to match the quality of the proteins used in the successful previous trials (5, 11). The other two hybrid proteins were purified from bacterial lysates to over 90% purity by size exclusion chromatography (SERP/41-3/HRPII) or by sequential cation and anion exchange and then size exclusion chromatography (SERP/ HRPII) (data not shown). The final products were dialyzed against phosphate-buffered saline–3 M urea and adjusted to 100 μg of protein per ml. Efficacy was tested following an experimental protocol identical to the one used in the previous successful trial (11).

Fifteen laboratory-raised Aotus azarae boliviensis karyotype VI monkeys were randomly assigned to one of four experimental groups (three groups of four and one group of three monkeys) and immunized with 1 ml of antigen or with the diluent alone (control group), both mixed with 100 μl of polyalphaolefin (4) as an adjuvant, on days 0, 21, and 42. Each vaccine dose was administered subcutaneously at two separate sites in the right and left flank and was well tolerated. The seroconversion results, as measured by enzyme-linked immunosorbent assay with SERP/HRPII as the solid-phase antigen and peroxidase-labelled rabbit anti-human immunoglobulin G (1:10,000 dilution; Pierce) as the secondary antibody, are shown in Fig. 1. All experimental monkeys developed comparable antibody responses to SERP/HRPII, irrespective of the immunogen. Control monkey sera did not react significantly (not shown). A boosting effect is obvious after the second injection in all three groups (Fig. 1), as well as after the third SERP/41-3/HRPII injection (Fig. 1C). This is similar to the seroconversion pattern observed previously (11). Prechallenge sera were also tested by immunofluorescence (IFA) for reactivity with P. falciparum schizonts. All preimmune sera and control group immune sera were negative (1:100 dilution). IFA titers from the experimental animals were all 1:1,600, except for animals A381 and A462 (titer, 1:800) and A452 and A292 (titer, 1:3,200). Thus, antibodies specific for native parasite determinants were induced. The relatively low IFA titers were comparable to those obtained in previous successful trials (5, 11).

FIG. 1.

Development of antibody responses in Aotus monkeys during the immunization period as determined by enzyme-linked immunosorbent assay. Monkeys in different immunization groups were immunized at weeks 0, 3, and 6 (indicated by arrows) and challenged at week 8 (indicated by an asterisk). All sera were tested for reactivity with SERP/HRPII in a 1:100 dilution. The hybrid antigens used for immunization were SERP/MSAI/HRPII (A), SERP/HRPII (B), and SERP/41-3/HRPII (C). Sera of the three control monkeys remained negative in this assay (not shown). OD, optical density.

At week 7 all monkeys were splenectomized, and at week 8 they received intravenously 2 × 10⁶ parasitized erythrocytes, which had been isolated from an Aotus monkey infected with an in vivo-passaged FUP-Cayenne isolate of P. falciparum (a kind gift of W. E. Collins) (Fig. 1). Monkey A293 appeared to have no spleen, although there was no prior history of splenectomy. The immunoglobulin G response of monkey A293 was nevertheless comparable to that of the other animals (Fig. 1B). Figure 2 shows the course of parasitemia after challenge. Two of three control animals rapidly developed a parasitemia which required mefloquine therapy (20 mg/kg of body weight orally) when parasitemia reached 10% (day 8 for A371 and day 10 for A320). In A432, parasitemia developed to 8.6% (day 10) and then fluctuated until rapidly reaching 19% (day 21), at which point mefloquine was administered (Fig. 2A). None of the three immunized groups showed a solid protective response (Fig. 2B to D). A340 (SERP/HRPII group) (Fig. 2C) and A292 (SERP/41-3/HRPII group) (Fig. 2D) showed low fluctuating parasitemias with a peak around 2.5% at the end of the observation period (day 25). Otherwise, parasitemias of the experimental animals did not significantly differ from those of the controls. No obvious correlation between prechallenge antibody levels and protection was evident.

FIG. 2.

Course of infection with the FUP-Cayenne isolate of P. falciparum in control Aotus monkeys (A) and in Aotus monkeys immunized with SERP/MSAI/HRPII (B), SERP/HRPII (C), and SERP/41-3/HRPII (D). Parasitemias of ≥10% were cured with mefloquine.

It seems unlikely that small conservative changes designed to improve SERP/HRPII and SERP/MSAI/HRPII expression in E. coli and to remove nonrelevant sequences adversely affected immune response development. After challenge, parasitemia developed markedly faster than in the previous trial, which had shown protection. Challenge with 2 × 10⁶ parasitized erythrocytes now resulted in high parasitemias on days 7 to 9 in 2 of the 3 controls and in 6 of the 12 experimental animals, whereas previously controls were untreated until day 14 (11). Also, one control and three experimental animals suffered recrudescence, which was not seen previously (11) or with later infections with the same parasite stock (5). It is remarkable that this apparent enhanced pathogenicity developed after a single passage in A. nancymai just before the present trial started. It is likely that this unintended pathogenicity influenced the experimental outcome. The protection of two monkeys in the SERP/HRPII and SERP/41-3/HRPII group may, however, reveal some protective effect of these vaccine candidates.

For demonstration of the protective potential of antigens in the primate model the pathogenicity of the challenge strain in the respective primate (sub)species, i.e., the equilibrium of immune response and pathogenicity, seems to be crucial (2, 12). The disturbance of this equilibrium may explain the discrepancy between previous successful trials (5, 11) and the present study. Recombinant proteins shown to be protective in the Aotus model (5, 11) failed to protect Saimiri monkeys, in which the course of parasitemia is quite different from that observed in Aotus monkeys. Similarly, no protection could be demonstrated in A. nancymai against the same challenge strain as was used in the successful trials with A. azarae boliviensis and A. lemurinus griseimembra (7a). The poor standardization of these models due to the scarcity of monkeys susceptible to human malaria remains an obstacle for the evaluation of human malaria vaccine candidates.