Transmission-blocking activity induced by malaria vaccine candidates Pfs25/Pvs25 is a direct and predictable function of antibody titer

Abstract

Background: Mosquito stage malaria vaccines are designed to induce an immune response in the human host that will block the parasite's growth in the mosquito and consequently block transmission of the parasite. A mosquito membrane-feeding assay (MFA) is used to test transmission-blocking activity (TBA), but in this technique cannot accommodate many samples. A clear understanding of the relationship between antibody levels and TBA may allow ELISA determinations to be used to predict TBA and assist in planning vaccine development.

Methods: Rabbit anti-Pfs25 sera and monkey anti-Pvs25 sera were generated and the antibody titers were determined by a standardized ELISA. The biological activity of the same sera was tested by MFA using Plasmodium gametocytes (cultured Plasmodium falciparum or Plasmodium vivax from malaria patients) and Anopheles mosquitoes.

Results: Anti-Pfs25 and anti-Pvs25 sera showed that ELISA antibody units correlate with the percent reduction in the oocyst density per mosquito (Spearman Rank correlations: 0.934 and 0.616, respectively), and fit a hyperbolic curve when percent reduction in oocyst density is plotted against antibody units of the tested sample. Antibody levels also correlated with the number of mosquitoes that failed to become infected, and this proportion can be calculated from the reduction in oocyst numbers and the distribution of oocysts per infected mosquito in control group.

Conclusion: ELISA data may be used as a surrogate for the MFA to evaluate transmission-blocking vaccine efficacy. This will facilitate the evaluation of transmission-blocking vaccines and implementation of this malaria control strategy.

Figure 1

Concentration-dependent TB activity of anti-Pfs25 monoclonal antibody 4B7. MAb 4B7 was serially diluted with pooled human O serum to give the antibody concentration in the feed as indicated (X-axis), mixed with mature in vitro-cultured P. falciparum gametocytes and fed to A. stephensi mosquitoes. Approximately one week later, oocysts were counted and % inhibition of oocyst density per mosquito relative to mosquitoes fed with a pooled human O serum is plotted (Y-axis). Line represents regression of result by use of a hyperbolic equation.

Figure 2

Kinetics of induction of rabbit anti-Pfs25 antibody. Groups of 4 rabbits were immunized with (i) 80 μg of Pfs25 adsorbed onto alum (blank bars), or (ii) 80 μg of Pfs25 formulated with Montanide ISA720 (hatched bars). Animals were immunized on days 0, 28 and 56, and bled on days 0, 28, 42 and 70. Sera were tested for anti-Pfs25 antibody concentration by ELISA relative to a standard rabbit anti-Pfs25 serum which has 100,000 units. ELISA values of individual rabbits are plotted and the geometric mean ± SE for the group are also shown. On day 70, the antibody units of the Pfs25-ISA group are significantly higher than those of the Pfs25-alum group (Mann-Whitney U-test, p < 0.0001).

Figure 3

Correlation of TB activity and antibody units of rabbit anti-Pfs25 sera. Groups of 4 rabbits were immunized with 80 μg of Pfs25 as described in the legend to Figure 2. Individual and pooled sera were diluted 1:2 and 1:8, and then tested for TB activity in mosquito membrane feeding assays using sexual stages of P. falciparum and A. stephensi mosquitoes. Anti-Pfs25 antibody units of each sample prior to dilution were determined by ELISA. Antibody unit values for the diluted sera are plotted on the X-axis and % inhibition of oocyst density per mosquito (open triangles) of the same sample or % inhibition of infected mosquitoes (closed circles) of the same sample is plotted on the Y-axis. There is a strong correlation between the antibody units and % inhibition of oocyst density per mosquito or % inhibition of infected mosquitoes (the Spearman Rank Correlations are 0.934 and 0.900, respectively). Lines represent regression of result by use of a hyperbolic equation: Solid line – correlation between antibody units and % inhibition of oocyst density per mosquito; broken line – correlation between antibody units and % inhibition of infected mosquitoes.

Figure 4

Effect of adjuvant selection and time after immunization on TB activity. Individual and pooled rabbit anti-Pfs25 sera were diluted 1:2 and 1:8, and then tested for TB activity in mosquito membrane feeding assays as described in the legend to Figure 3. (a) data were categorized according to the adjuvant used for immunization (open circles: alum; closed circles: Montanide ISA720) while in (b) data were categorized according to number of days after initial immunization (open triangles: day 42; closed triangles: day 70). Lines represent regression of result by use of a hyperbolic equation. Four data points, which showed more than 30,000 antibody units and 100% inhibition of oocyst density per mosquito, are not shown in these figures.

Figure 5

Kinetics of antibody response to immunization with Pvs25 in rhesus monkeys. Groups of 5 monkeys were immunized with (i) 15 μg of Pvs25 adsorbed onto 600 μg of alum (blank bars), or (ii) 15 μg of Pvs25 adsorbed onto 600 μg of alum and 250 μg of CpG 10105 (hatched bars). Animals were immunized on days 0, 28 and 181, and bled on days 0, 42, 90 and 195. All sera were tested by ELISA and compared to a standard rhesus anti-Pvs25 serum which has 20,000 units. ELISA values of individual monkeys are plotted and the geometric mean ± SE for the group are also shown.

Figure 6

Correlation of TB activity with antibody titer using anti-Pfs25 sera from rabbits and anti-Pvs25 sera from rhesus monkeys. Rabbits were immunized with Pfs25 and the antisera were tested in mosquito membrane feeding assays with P. falciparum parasites and A. stephensi mosquitoes as described in legend to Figure 3. Rhesus monkeys were immunized with Pvs25 as described in legend to Figure 5 and the antiserum was diluted with a pool of normal human AB+ serum as 1:2 or 1:8 dilutions. The diluted monkey sera were tested in membrane feeding assays using parasites derived from vivax-infected patients in Thailand and A. dirus mosquitoes. Antibody units in the sera were determined by standardized ELISA. (a): Rabbit data (open circles) and individual monkey data (closed circles) are shown. There is a significant correlation between antibody units and % inhibition of oocyst density per mosquito in rabbit and in monkey sera (the Spearman Rank Correlations are 0.934 and 0.616, respectively). Lines represent regression of monkey result by use of a hyperbolic equation. Four points of rabbit data, which showed more than 30,000 antibody units and 100% inhibition of oocysts, are not shown in this figure. (b):The values of the test sera were grouped as shown. Average percent oocyst inhibition for rabbit sera (blank bars) or monkey sera (hatched bars) in each ELISA range + SD are shown.

Figure 7

Comparison of predicted and measured percentage of infected mosquitoes. Predicted values are the average of 40 computer simulations per point for membrane feeds based on the experimental feeds measuring TB activity of rabbit anti-Pfs25 sera (a) and rhesus anti-Pvs25 sera (b). These are the same feeds shown in Figure 4 and Figure 6, respectively.

Figure 8

Comparison of predicted and measured oocyst density per mosquitoes. Experimental data (open circles) are shown for anti-Pfs25 (a) and anti-Pvs25 (b) and are the same feeds shown in Figure 4 and Figure 6, respectively. A single simulated data set (closed circles) is shown for each.