A new double-antibody sandwich ELISA targeting Plasmodium falciparum aldolase to evaluate anti-malarial drug sensitivity

Background The standard in vitro test to assess anti-malarial activity of chemical compounds is the [3H]hypoxanthine incorporation assay. It is a radioactivity-based method to measure DNA replication of Plasmodium in red blood cells. The method is highly reproducible, however, the handling of radioactive material is costly, hazardous and requires the availability of appropriate technology and trained staff. Several other ways to evaluate in vitro anti-malarial activity do exist, all with their own assets and limitations. Methods The newly developed double-antibody sandwich ELISA described here is based on the properties of a non-overlapping pair of monoclonal antibodies directed against Plasmodium falciparum aldolase. This glycolytic enzyme possesses some unique nucleotide sequences compared to the human isoenzymes and has been highly conserved through evolution. Out of twenty possibilities, the most sensitive antibody pair was selected and used to quantitatively detect parasite aldolase in infected blood lysates. Results A total of 34 compounds with anti-malarial activity were tested side-by-side by ELISA and the [3H]hypoxanthine incorporation assay. The novel ELISA provided IC50s closely paralleling those from the radioactivity-based assay (R = 0.99, p < 0.001). At the investigated assay conditions (72 h incubation time, parasitaemia = 0.3%), the assay was found to be reproducible and easy to perform. Conclusion The newly developed ELISA presents several advantages over the comparative method, the [3H]hypoxanthine incorporation assay. The assay is highly reproducible, less hazardous (involves no radioactivity) and requires little and cheap technical equipment. Relatively unskilled personnel can conduct this user-friendly assay. All this makes it attractive to be employed in resource-poor laboratories.


Background
Several techniques exist to measure anti-malarial activity of chemical compounds. The most commonly used method, especially in well-equipped laboratories, is the [ 3 H]hypoxanthine incorporation assay [1]. This method is highly reproducible, however, the handling of radioactive material is costly, hazardous and quite complex and, therefore, problematic for resource-poor locations. Moreover, radioactive material is not uniformly authorized worldwide, limiting its application geographically. A low-cost alternative is the schizont maturation assay, standardized by the World Health Organization. However, this test can only be carried out by the experienced microscopist, is very labour-intensive and prone to individual variability. A method that is simple to establish, highly reproducible, that requires little technical equipment and could be applicable to a field laboratory, is the enzyme-linked immunosorbent assay (ELISA). A few commercialized ELISA tests are already available, targeting either Plasmodium falciparum lactate dehydrogenase (pLDH) or histidine-rich protein 2 (HRP2) [2][3][4][5][6]. Like P. falciparum aldolase, pLDH is very much conserved between P. falciparum isolates [7]. It presents some unique differences to the human LDH, and its level can be used to determine the drug susceptibility of malaria parasites [8]. The existing commercial kits testing pLDH are suitable with an initial parasitaemia of 0.005% of cultivated or natural strains [5]. HRP2 has recently been reported to show extensive protein sequence diversity (mainly insertions) in all of the analysed 75 P. falciparum isolates collected from geographically different areas [9]. Importantly, it was demonstrated that the HRP2 protein diversity had an effect on the sensitivities of the HRP2 detection antibodies. The same group also reports that the aldolase protein sequence shows no insertions by analysing 36 of the original 75 P. falciparum isolates [10]. This prompted us to develop a suitable double-antibody sandwich ELISA detecting P. falciparum aldolase to evaluate anti-malarial drug sensitivity. The newly developed aldolase ELISA was compared to the [ 3 H]hypoxanthine incorporation assay, testing anti-malarial compounds such as OZ277 [11,12], artesunate (AS), chloroquine (CQ), pyrimethamine (PYR) and mefloquine (MEF).

Chemicals and materials
OZ277 tosylate (MW: 565) was provided by J.L. Vennerstrom (Nebraska, USA), pyrimethamine (MW: 249) and mefloquine hydrochloride (MW: 415) were gifts from F. Hoffmann-LaRoche (Basel, Switzerland), artesunate (MW: 384) was donated by Guilin Pharma Corp. (Guilin Guangxi, China) and chloroquine diphosphate (MW: 516) was purchased from Sigma. Further anti-malarial compounds were obtained from the NGBS malaria pro-gramme, a consortium formed by the Novartis Institute for Tropical Diseases, the Genomics Institute of the Novartis Research Foundation, the Biomedical Primate Research Center and the Swiss Tropical Institute (compounds were provided by M. Rottmann, Swiss Tropical Institute Basel, Switzerland). All anti-malarial compounds were dissolved in dimethylsulfoxide (DMSO) at 10 mg/ ml, except chloroquine, which was dissolved in water. The stock solutions were kept at 4°C for a maximum of six months. Dilutions were prepared from the stock solution in hypoxanthine-free culture medium immediately before use. [ 3 H]hypoxanthine was purchased from Amersham Bioscience (Buckinghamshire, UK).

In vitro growth-inhibition assay
Plasmodium falciparum (NF54) growth was assessed by measuring the incorporation of the nucleic acid precursor [ 3 H]hypoxanthine [18]. IC 50 values were earlier found to be 0.91 ± 0.12 ng/ml for OZ277, 1.6 ± 0.1 for artesunate, 5.1 ± 0.8 for chloroquine, 5.8 ± 0.2 for mefloquine [11] and 5.6 ± 0.5 ng/ml for pyrimethamine [19]. Infected erythrocytes (100 μl per well with 2.5% haematocrit and 0.3% parasitaemia) were added to each drug titrated in 100 μl duplicates over a 64-fold range. After 48 h incubation, 0.5 μCi of [ 3 H]hypoxanthine in 50 μl medium was added and plates were incubated for an additional 24 h. Parasites were harvested onto glass-fiber filters and radio-activity was counted using a Betaplate liquid scintillation counter (Wallac, Zurich). The results were recorded as counts per minute (cpm) per well at each drug concentration and expressed as a percentage of the untreated controls.
Double antibody sandwich ELISA 96-well microtiter plates (Maxisorp, Nunc) were coated with 100 μl/well of the capture antibody P.41-24/10-11 (5 μg/ml in PBS) overnight at 4°C. The plates were then saturated with 200 μl/well of 1% BSA in PBS-Tween20 (0.05%) for 1 hour at 37°C. 50 μl of the test samples were then added and incubated 1 h at 37°C. Then, 50 μl/well of the detection antibody P.41-2/3-7 -HRP (0.8 μg/ml) were added, and the plates incubated 1 h in the dark at room temperature. Finally, 100 μl/well of the substrate (Tetramethylbenzidine) solution was added and the reac-tion stopped after 5 min with 100 μl 1 N H 2 SO 4 . The optical densities (ODs) were measured at 450 nm by an ELISA plate reader (Biotrak visible plate reader, Amersham Pharmacia Biotech). Between each incubation step, the plates were washed four times with PBS-Tween (0.05% Tween20).

Statistics
Fifty percent inhibitory concentrations (IC 50 s) were determined in XLfit from IDBS by a 4 parameter logistic model. After controlling the normal distribution of the data, a correlation coefficient between the results obtained sideby-side by the two methods was determined by a standard Pearson's correlation analysis (Stata 9.1) and a Bland-Altman plot was performed, testing the agreement between them (Stata 9.1).

Results
Five anti-recombinant aldolase monoclonal antibodies were tested both as capture as well as detection antibodies (HRP-labeled) in a double-antibody sandwich ELISA targeting P. falciparum recombinant aldolase P41/7 (highest concentration was 10 μg/ml). Ultimately, the antibody pair P.41-24/10-11; P.41-2/3-7+HRP was selected, showing a good specificity (steep curve) and the broadest dynamic range (OD span An excellent correlation coefficient of R = 0.991 (P < 0.001) was found when additional 29 anti-malarial compounds were tested and both log IC 50 s of all compounds, including OZ277, AS, CQ, PY and MEF, plotted ( Figure  1).
To test the agreement between ELISA and [ 3 H]hypoxanthine incorporation assay, the difference in log IC 50 s obtained from the two methods were plotted against their mean value in a Bland-Altman plot ( Figure 2). The two methods were found to agree, as > 96% of the values lie in acceptable broad limits of agreement (-0.85 to 0.32).

Discussion
The novel ELISA provides IC 50 s closely paralleling those from the standard [ 3 H]hypoxanthine incorporation assay (Table 1), with P.41-24/10-11 as capture antibody and P.41-2/3-7-HRP as detection antibody targeting P. falciparum aldolase. The excellent correlation between IC 50 s determined by the [ 3 H]hypoxanthine incorporation assay and by aldolase ELISA (R = 0.991, P < 0.001; Figure 1) was further confirmed by the Bland-Altman plot (Figure 2), showing a general trend of the two methods to agree nicely. The limits of agreement correlate with the findings of Noedl et al who compared the [ 3 H]hypoxanthine incorporation assay and the HRP2 ELISA [20]. Interestingly, the aldolase ELISA tends in all cases to measure slightly higher IC 50 s than the [ 3 H]hypoxanthine incorporation assay. However, this observation should have no impact on the power of the assay.
Aldolase is commonly used as a pan-malaria antigen in rapid diagnostic tests (RDTs) [21]. It can, therefore, not be ruled out that our monoclonal antibodies could crossreact with aldolase from other Plasmodium species. This should be kept in mind when embarking in future drug susceptibility testing efforts in species-overlapping zones.
Aldolase-and pLDH-based RDTs are reported to be less sensitive than HRP2-based tests, due to the transient presence in blood of the first [21,22]. However, since the lowest parasitaemia tested in this study (0.03%) did result in a solid signal to background ratio of ~5, the sensitivity is not expected to be an issue for drug sensitivity testing in the field. Furthermore, the aldolase amino acid sequence is highly conserved, preventing detection failures, as seen in the case of HRP2 [9]. The aldolase ELISA presents no drawbacks or assets over pLDH-based assays.

Conclusion
The novel aldolase ELISA assay is highly reproducible, less hazardous than the [ 3 H]hypoxanthine incorporation assay and requires little and cheap technical equipment.
Relatively unskilled personnel can conduct this userfriendly assay. All this makes it attractive to be employed in resource-poor laboratories.
The fact that the P. falciparum aldolase has been highly conserved during evolution [10] renders aldolase an antibody target of choice for the analysis of field isolates. It  also provides a considerable advantage over the HRP2 ELISA, since the HRP2 gene sequence is showing an extensive diversity, altering the sensitivity of some commercial diagnostic tests.
However, before proceeding to drug susceptibility testing in endemic countries, further investigations with more parasite isolates should be performed in order to determine the lowest measurable parasitaemia (sensitivity) of the aldolase ELISA as well as if the here described monoclonal antibodies cross-react with aldolase from other Plasmodium species.
Difference in log IC 50 s obtained from the the aldolase ELISA and the [ 3 H]hypoxanthine method